1: \documentclass[12pt,preprint]{emulateapj}
2: \slugcomment{{\sc Accepted to ApJS:} April 23, 2008}
3: \usepackage{lscape}
4:
5: \newcommand{\mapiii}{MAPPINGS\,\textsc{iii}}
6: \newcommand{\mapii}{MAPPINGS\,\textsc{ii}}
7: \newcommand{\kms}{\ifmmode {\rm\,km~s^{-1}} \else ${\rm\,km\,s^{-1}}$\fi}
8: \slugcomment{}
9: \shorttitle{The MAPPINGS {\sc III} Shock Model Library}
10: \shortauthors{Allen et al.}
11:
12: \begin{document}
13:
14: \title{The MAPPINGS {\sc III} Library of Fast Radiative Shock Models}
15:
16: \author{Mark G. Allen }
17: \affil{Observatoire de Strasbourg UMR 7550
18: Strasbourg 67000, France}
19: \email{allen@astro.u-strasbg.fr}
20:
21: \author{Brent A. Groves}
22: \affil{Sterrewacht Leiden, Leiden University, Neils Bohrweg 2, Leiden 2333-CA
23: The Netherlands}
24: \email{brent@strw.leidenuniv.nl}
25:
26: \author{Michael A. Dopita }
27: \affil{Research School of Astronomy and Astrophysics,
28: Australian National University, Cotter Road , Weston Creek ACT 2611,
29: Australia }
30:
31: \author{Ralph S. Sutherland }
32: \affil{Research School of Astronomy and Astrophysics,
33: Australian National University, Cotter Road , Weston Creek ACT 2611,
34: Australia }
35:
36: \author{Lisa J. Kewley}
37: \affil{Institute for Astronomy, University of Hawaii, 2680 Woodlawn Drive
38: Manoa, HI 96822. USA}
39:
40:
41: \begin{abstract}
42: We present a new library of fully-radiative shock models calculated with the
43: \mapiii\ shock and photoionization code. The library consists of grids of
44: models with shock velocities in the range $v_{s}$=100-1000\ km\ s$^{-1}$ and
45: magnetic parameters B/$\sqrt{\rm n}$ of 10$^{-4}$-10 $\mu$G\ cm$^{3/2}$ for
46: five different atomic abundance sets, and for a pre-shock density of 1.0\
47: cm$^{-3}$. Additionally, Solar abundance model grids have been calculated for
48: densities of 0.01, 0.1, 10, 100, and 1000\ cm$^{-3}$ with the same range in
49: $v_{s}$ and B/$\sqrt{\rm n}$. Each model includes
50: components of both the radiative shock and its photoionized precursor, ionized
51: by the EUV and soft X-ray radiation generated in the radiative gas. We
52: present the details of the ionization structure, the column densities,
53: and the luminosities of the shock and its precursor. Emission line ratio predictions
54: are separately given for the shock and its precursor as well as for the
55: composite shock+precursor structure to facilitate comparison with observations
56: in cases where the shock and its precursor are not resolved. Emission line
57: ratio grids for shock and shock+precursor are presented on standard line ratio
58: diagnostic diagrams, and we compare these grids to observations of radio
59: galaxies and a sample of AGN and star forming galaxies from the Sloan Digital
60: Sky Survey.
61: This library is available online, along with a suite of tools to
62: enable the analysis of the shocks and the easy creation of emission
63: line ratio diagnostic diagrams. These models represent a significant increase in
64: parameter space coverage over previously available models, and
65: therefore provide a unique tool in the diagnosis of emission by shocks.
66: \end{abstract}
67:
68:
69: \keywords{ hydrodynamics - shock waves - ISM: abundances,- Galaxies: Nuclei, Galaxies:
70: Seyfert - infrared: ISM, Ultraviolet: ISM, X-rays: ISM}
71:
72:
73: \section{Introduction}
74:
75: Supersonic motions are a common phenomenon in our complex and fascinating
76: Universe. The kinetic energy of such motions will almost inevitably be
77: eventually dissipated through radiative shocks. Cloud-cloud collisions, the
78: expansion of \ion{H}{2} regions into the surrounding interstellar medium,
79: outflows from young stellar objects, supernova blast waves, outflows from
80: active and starburst galaxies, and collisions between galaxies are all
81: examples of astrophysical situations in which radiative shock waves provide an
82: important component of the total energy budget and may determine the line
83: emission spectrum.
84:
85: In this paper we do not consider very slow (molecular) shocks or indeed, the
86: faster atomic shocks for which the theory has been well-developed by such
87: authors as \citet{Dopita77} or \citet{Raymond79}. Rather, we consider only
88: fast shocks, where the ionizing radiation generated by the cooling of hot
89: gas behind the shock front generates a strong radiation field of extreme
90: ultraviolet and soft X-ray photons, which leads to significant photoionizing
91: effects. The detailed theory of steady-flow photoionizing (or auto-ionizing)
92: shocks was developed by \cite{sutherland1993} and \cite{dopita1995,dopita1996}
93: (hereafter DS95 and DS96). A detailed text-book development of the theory
94: of shocks is given in \cite{dopita2003}.
95:
96:
97: \begin{deluxetable}{lrrrrr}
98: \tablecolumns{6}
99: \tablewidth{0pc}
100: \tablecaption{Abundance Sets (by number wrt Hydrogen) \label{abund_table}}
101: \tablehead{
102: \colhead{Element} &
103: \colhead{Solar} &
104: \colhead{Solar $\times$2} &
105: \colhead{dopita2005} &
106: \colhead{LMC} &
107: \colhead{SMC} }
108: \startdata
109: H & 0.00 & 0.00 & 0.00 & 0.00 & 0.00 \\
110: He & -1.01 & -1.01 & -1.01 & -1.05 & -1.09 \\
111: C & -3.44 & -3.14 & -4.11 & -3.96 & -4.24 \\
112: N & -3.95 & -3.65 & -4.42 & -4.86 & -5.37 \\
113: O & -3.07 & -2.77 & -3.56 & -3.65 & -3.97 \\
114: Ne & -3.91 & -3.61 & -3.91 & -4.39 & -4.73 \\
115: Na & & & -6.35 & -4.85 & -5.92 \\
116: Mg & -4.42 & -4.12 & -5.12 & -4.53 & -5.01 \\
117: Al & -5.53 & -5.23 & -7.31 & -4.28 & -5.60 \\
118: Si & -4.45 & -4.15 & -5.49 & -5.29 & -4.69 \\
119: S & -4.79 & -4.49 & -5.01 & -7.23 & -5.41 \\
120: Cl & & & -6.70 & & -7.30 \\
121: Ar & -5.44 & -5.14 & -5.44 & -5.71 & -6.29 \\
122: Ca & -5.88 & -5.58 & -8.16 & -6.03 & -6.16 \\
123: Fe & -4.63 & -4.33 & -6.55 & -4.77 & -5.11 \\
124: Ni & & & -7.08 & -6.04 & -6.14 \\
125: \enddata
126: \end{deluxetable}
127:
128: In photoionizing shocks, the flux of the ionizing radiation emitted
129: by the shock increases in approximate proportion to the energy flux through the
130: shock ($\propto v_{\rm s}^{3}$). The ratio of this flux in advance of
131: the shock to the
132: pre-shock density, classified as the ionization parameter, determines the
133: velocity of the photoionization front that is driven into the pre-shock gas. At
134: low values of the ionization parameter, the velocity of the photoionization
135: front is lower than the velocity of the shock and the ionizing photons are
136: absorbed in the immediate vicinity of the shock front. The effect of this is
137: to change the ionization state of the gas feeding across the shock front. As
138: the velocity of the shock increases, the emitted ionizing flux, and therefore
139: the velocity of the photoionization front, increases rapidly. At shock
140: velocities above a certain limit ($v_{\rm s} \approx$ 170 km\ s$^{-1}$), the
141: ionization front velocity exceeds that of the shock (and is supersonic
142: with respect to the pre-shock gas) and the photoionization
143: front detaches from the shock front as an R-Type ionization
144: front\footnote{ R-Type ionization fronts are defined as being supersonic with
145: respect to the gas ahead of the front. (subsonic ionization fronts are
146: classified as D-Type) see \citet{mckee1980}}. This front expands to form a
147: precursor \ion{H}{2} region ahead of the shock. At the highest shock
148: velocities, the photoionized precursor emission may come to dominate the
149: optical emission of the shock, and the global radiative shock spectrum
150: provides a rich mixture of emission lines from both high- and low-ionization
151: species.
152:
153: \begin{figure}[t]
154: \includegraphics[scale=0.5]{f1.ps}
155: \caption{Ionizing spectra generated by $v_s$=250, 500, 750 and 1000
156: km\ s$^{-1}$ shock models
157: with n=1.0 cm$^{-3}$ and solar abundance. \label{specfig1}}
158: \end{figure}
159:
160:
161: \begin{figure}[t]
162: \includegraphics[scale=0.5]{f2.ps}
163: \caption{Ionizing spectra of $v_s$=400 and 1000 km\ s$^{-1}$ shock
164: models shown on $\nu F_{\nu}$ scale. \label{specfig2}}
165: \end{figure}
166:
167: Appreciable photoionization effects are also produced in the gas behind the
168: shock front, near to the recombination region of the shock. However, in this
169: shock region the velocity of the ionization front is much slower due to the
170: gas compression through the shock front and the ionization front stays trapped
171: in the recombination region to much higher shock velocities. When magnetic
172: effects are negligible, the compression of the gas in the shocked region is
173: proportional to the square of the Mach number in the pre-shock gas, ${\cal M}
174: = v_{\rm s}/c_{\rm II}$, where $c_{\rm II}$ is the sound speed in the
175: pre-shock gas. When the magnetic field pressure in the post-shock gas
176: dominates over the gas pressure, the compression factor is determined by the
177: Alfv\'en Mach Number, ${\cal M}_{\rm A}$, which is the ratio of the shock
178: velocity to the Alfv\'en velocity; $v_{\rm A} = ({\rm B}^2 /4\pi \rho)^{1/2}$,
179: where B is the transverse component of the pre-shock magnetic field, and
180: $\rho$ is the pre-shock mass density. Thus the presence of magnetic
181: fields act to limit the compression through the shock. In order to account for
182: this effect in computations of the shock structure, DS96 developed the concept
183: of a magnetic parameter; B/$\sqrt{\rm n}$, where n is the pre-shock particle
184: number density.
185:
186: \begin{figure*}
187: \centering
188: \includegraphics[scale=0.80]{f3_col.ps}[t]
189: \caption{The hydrogen ionization structure, temperature profile and density
190: profile of a shock and precursor. The vertical axis represents i) the
191: ionization fraction X$_{ion}$ (values $\le 1$) for the ionization structure
192: curves, ii) the electron temperature profile (T$_e$) in degrees Kelvin, shown
193: as the upper line plotted in the figure, and iii) the hydrogen density
194: (n$_{\rm H}$) in cm$^{-3}$, shown by the dotted line in the shock region
195: (right) panel of the figure. As n$_{\rm H}$ is constant in the precursor
196: (left) it is not plotted. The horizontal axis represents the time since
197: passage of the shock front, with positive values for the shock structure shown
198: in the right of the figure, and negative values for the precursor region where
199: the shock is yet to arrive. The numerical labels indicated on the temperature
200: profile of the shock correspond to different regions of the shock structure as
201: described in the text. The model plotted here is our fiducial model with
202: solar abundances, equipartition magnetic field, a precursor density of n$_{\rm
203: H}=1$ cm$^{-3}$, and shock velocity of $v_{\rm s}$=500 km\ s$^{-1}$, as
204: labelled in the upper right corner of the figure. See the electronic edition
205: of the Journal for a color version of this figure.
206: \label{Hstruct_eg}}
207: \end{figure*}
208:
209: The grid of low density photoionizing shock models described in DS96 has
210: proved to be a valuable resource to the astronomical community for assessing
211: the role of photoionizing shocks in a range of astrophysical objects. This
212: utility is increased by their availability in electronic form from the AAS
213: CD-ROM Series, Vol 7. \citep[see][]{leitherer1996}. These models have most
214: often been applied in studies of the narrow line regions (NLR) of active
215: galaxies, developing the debate on the relative roles of shock and/or
216: photoionization excitation of the NLR, and the question of the radiative
217: versus mechanical energy output of active galactic nuclei (AGN). Other uses of
218: the models have included studies of ionized gas around high velocity clouds
219: \citep{fox2004}, and the intergalactic medium \citep{shull2003}.
220:
221:
222: In DS95 these models were applied to optical line ratios observed in
223: NLRs. They found that Seyferts have values close to the shock + precursor
224: predictions and that Low Ionization Emission Line Region (LINER) galaxies fall within
225: the range of the shock-only models. In
226: \cite{allen1998} we highlighted the importance of UV line ratios which help
227: separate shock and photoionization models, and we defined a set of UV line
228: ratio diagnostic diagrams. These UV diagnostics were applied to individual
229: objects M~87 \citep{dopita1997}, NGC~5728 and NGC~5643 \citep{evans1999} and
230: NGC~1068 \citep{groves2004}. NGC~2992 was also investigated in detail with
231: optical diagnostics and other multi-wavelength data \citep{allen1999}.
232:
233: The DS95 model grid has also been extensively used by other authors in studies
234: of AGN. \cite{best2000} and \cite{Inskip02} used optical and UV line ratios to
235: show that shocks associated with radio sources in $z\sim1$ 3CR radio galaxies
236: can dominate the kinematics and ionization as the radio source expands through
237: the interstellar medium of the host galaxy. These models have also application
238: to the analysis of emission lines in high redshift radio galaxies
239: \citep{villar1997, maxfield2002, debreuck2000, villar2003, iwamuro2003, Reuland07}, and in
240: studies of individual low redshift active galaxies including NGC~4151
241: \citep{nelson2000}, M~51 \citep{bradley2004}, M~87 \citep{sabra2003}, Mkn~78
242: \citep{whittle2005}, 3C~299 \citep{feinstein1999}, NGC~2110 and NGC~5929
243: \citep{ferruit1999}.
244:
245: Studies of individual objects make it obvious that higher dimensional modeling
246: of the physical structures and radiation fields associated with shocks is
247: necessary in order to be able to more closely model these complex physical
248: systems, and to be able to draw deeper conclusions. Indeed the 1-D,
249: steady-flow nature of these models is acknowledged as their greatest
250: limitation, since all fast shocks are subject to thermal instability, and the
251: generation of secondary shocks within the cooling zone. To address these
252: issues, much effort is being put into higher dimensional numerical simulations
253: of radiative shocks, in particular in the area of understanding the role of
254: local thermal and dynamical instabilities on both the shock structure and on
255: the emergent spectrum \citep{sutherland2003a}. However, given the
256: complexities inherent in their computation, and the large amount of
257: supercomputer time required, complete grids of model predictions from full 3-D
258: models are some time off, and until then, modeling will most likely favour
259: simulations of individual and specific cases.
260:
261: Thus, the simpler 1-D models remain a useful tool for comparing with
262: observations of both individual objects and in the investigation of the mean
263: parameters of groups of objects observed in surveys. Furthermore, these models
264: serve as a stepping stone towards higher dimensional models. For example,
265: under some simplifying assumptions the radiative properties of 1-D models may
266: be mapped onto those of 3-D hydrodynamical models. In particular, thermally
267: unstable shocks of a given velocity, $v_{s}$, behave similarly to steady-flow
268: shocks with shock velocities $\sim \frac{2}{3}v_{s}$ \citep{sutherland2003a}.
269: Bearing in mind these effects thermal instability, we
270: advise that care should be taken in the use of 1-D models to derive physical
271: parameters in individual cases, or at least allow for this factor of 2/3.
272:
273: In this paper we present a new library of 1-D steady flow photoionizing shock
274: models, calculated with an updated version of the shock modeling code,
275: \mapiii. The library is an improvement upon and extension of the DS95 \&
276: DS96 models, and includes a range of chemical abundance sets, pre-shock
277: densities from 0.01 to 1000 cm$^{-3}$, velocities up to 1000 km\ s$^{-1}$,
278: magnetic fields from 10$^{-10}$ to 10$^{-4}$ G, and magnetic parameters
279: (B/$\sqrt{\rm n}$) from 10$^{-4}$ to 100 $\mu$G\ cm$^{3/2}$. These new
280: model grids therefore
281: supersede the previous models, and are designed to be of maximum utility
282: to observers for comparing observations with photoionizing shock
283: models.
284:
285: As well as reiterating the technique used in MAPPINGS to create the
286: shock models, we discuss in detail the radiation fields generated by
287: shocks, and the resulting ionization, density and temperature
288: structure of shocks for the range of parameters considered.
289: We also present the resulting model grids on a range of ultraviolet,
290: optical and infrared emission-line diagnostic diagrams, discussing the
291: effects of the various parameters upon the grids, and demonstrate
292: their usefulness and applicability by comparing these grids with
293: emission-line galaxy sample from the Sloan Digital Sky Survey (SDSS).
294:
295: \begin{figure*}[htb]
296: \centering
297: \includegraphics[scale=0.7]{f4a_col.ps} \\
298: \includegraphics[scale=0.7]{f4b_col.ps} \\
299:
300: \caption{The hydrogen ionization structure, temperature profile and
301: density profile of a shock and precursor for
302: models with shock veocities of $v_{\rm s}$ = 200 and 1000
303: km\ s$^{-1}$ and equipartition magnetic field. The axes are as
304: described in Figure~\ref{Hstruct_eg}.
305: See the electronic edition of the Journal for a color version of this
306: figure.\label{Hstruct_V} }
307: \end{figure*}
308:
309: \begin{figure*}[htb]
310: \centering
311: \includegraphics[scale=0.7]{f5a_col.ps} \\
312: \includegraphics[scale=0.7]{f5b_col.ps} \\
313: \caption{The hydrogen ionization structure, temperature profile and
314: density profile of
315: models with different magnetic parameter. Models of $v_{\rm s} = 500$
316: km\ s$^{-1}$ are shown for the magnetic parameters of 0.0001 and 10
317: $\mu$G\ cm$^{3/2}$.
318: The axes are the same as for Figure~\ref{Hstruct_eg}. See the
319: electronic edition of the Journal for
320: a color version of this figure.
321: \label{Hstruct_B}}
322: \end{figure*}
323:
324:
325: We introduce the \mapiii\ online shock library, the
326: main access point for the results of the shock models, which includes
327: both the resulting emission lines and column depths from the models,
328: as well as the full results of the \mapiii\ models for detailed
329: analysis. Also included as part of the library are the analysis tools
330: used within this work, providing a simple and easy mechanism to
331: determine the shock diagnostic power of any emission-line ratios.
332:
333:
334: \section{Modeling Technique}
335:
336: The models presented here have been calculated with the \mapiii\ shock and
337: photoionization modeling code, version {\sc III}q. This is an updated version
338: of the \mapii\ code which was described in
339: \cite{sutherland1993b}. The main enhancements of version III include the use
340: of a higher resolution radiation vector, and a more stable scheme for choosing
341: time-steps which allows computation of higher velocity shocks. Other
342: improvements include the explicit inclusion of all the He~{\sc II} emission
343: lines, and a corrected computation of the neutral oxygen emission. Further details
344: on the use of and the chronology of the development of MAPPINGS models may be
345: found at the MAPPINGS Online web
346: pages\footnote{\url{http://www.ifa.hawaii.edu/\~{}kewley/Mappings}}. Various
347: improvements to the calculation of photoionization processes have been
348: implemented in version III \citep{groves2004a,groves2007} although the dust
349: effects discussed there have not been considered in these shock models, which
350: are fundamentally dust-free. This approximation is probably a physical one,
351: since in the faster shocks dust will be effectively destroyed by grain-grain
352: collisions, through both shattering and spallation, and by thermal
353: sputtering. The physics of these processes have been discussed in detail
354: \citet{Draine79, Dwek92, Jones94, Jones96, Pineau97} and \citet{perna2002}.
355: We have in
356: any case run models with depleted abundances to facilitate comparison with the
357: earlier DS96 high-velocity shock modeling, so our different model grids can be
358: used to provide some estimate of the effect of dust depletion on the output
359: spectra.
360:
361: \begin{figure*}[t]
362: \centering
363: \includegraphics[scale=0.6]{f6a_col.ps} \\
364: \includegraphics[scale=0.6]{f6b_col.ps} \\
365: \includegraphics[scale=0.6]{f6c_col.ps} \\
366: \caption{The hydrogen ionization structure, temperature profile and density
367: profile of models with different pre-shock density. Models of $v_{\rm s} =
368: 500$ km\ s$^{-1}$ and equipartition magnetic field are shown for pre-shock
369: densities of 0.1, 10 and 1000 cm$^{-3}$. The axes are the same as for
370: Figure~\ref{Hstruct_eg}. See the electronic edition of the Journal for a
371: color version of this figure.
372: \label{Hstruct_dens} }
373: \end{figure*}
374:
375:
376: \begin{figure*}[t]
377: \centering
378: \includegraphics[scale=0.6]{f7a_col.ps} \\
379: \includegraphics[scale=0.6]{f7b_col.ps} \\
380: \includegraphics[scale=0.6]{f7c_col.ps} \\
381: \caption{The hydrogen ionization structure, temperature profile and density
382: profile of models with different atomic abundances. Models of $v_{\rm s} =
383: 500$ km\ s$^{-1}$ and equipartition magnetic field are shown for the
384: 2$\times$solar, dopita2005 and SMC abundance sets. The axes are the
385: same as for Figure~\ref{Hstruct_eg}. See the electronic edition of
386: the Journal for a color version of this figure.
387: \label{Hstruct_abund}}
388: \end{figure*}
389:
390:
391:
392: \mapiii\ uses exactly the same computational recipe for calculating the flow
393: solution of these one-dimensional radiative shock models as described in
394: detail in DS96 for \mapii . Briefly, the hydrodynamics of the flow
395: are derived from the Rankine-Hugoniot jump conditions including the magnetic
396: field terms. The four Rankine-Hugoniot conditions, equations \ref{rh1} -
397: \ref{rh4}, represent the conservation of mass, conservation of momentum, the
398: condition that the magnetic field is locked into the ionized plasma, and
399: conservation of energy (see \cite{heng2007} for a generalization of these conditions).
400: These equations connect two points in the flow at
401: times $t$ and $t_0$ with density, velocity, pressure, internal energy, and
402: transverse magnetic field $\rho$, $V$,$P$, $U$, B, and $\rho_0$,
403: $V_0$,$P_0$, $U_0$, B$_0$, respectively. $\bar{\Lambda}$ is the mean cooling
404: rate of the plasma over the time step. As set out in DS96, by choosing the
405: subscripted flow variables as the point immediately in front of the shock, the
406: equations can be reduced to a quadratic expression for the flow velocity. By
407: choosing appropriate fractions of the geometric mean of the plasma timescales,
408: the physical conditions in the plasma can be followed smoothly.
409:
410: \begin{equation}
411: \label{rh1}
412: \rho v=\rho_0 v_0
413: \end{equation}
414:
415: \begin{equation}
416: \label{rh2}
417: \rho v^2 + P + \frac{{\rm B}^2}{8\pi} =\rho_0 v_0^2 + P_0 + \frac{{\rm
418: B}_0^2}{8\pi}
419: \end{equation}
420:
421: \begin{equation}
422: \label{rh3}
423: \frac{{\rm B}}{\rho} = \frac{{\rm B}_0}{\rho_0}
424: \end{equation}
425:
426: \begin{equation}
427: \label{rh4}
428: \frac{v^2}{2} + U + \frac{P}{\rho} + \frac{{\rm B}^2}{4\pi} +
429: \bar{\Lambda}(t - t_0)
430: = \frac{v_0^2}{2} + U_0 + \frac{P_0}{\rho_0} + \frac{{\rm B}_0^2}{4\pi}
431: \end{equation}
432:
433: The main calculation to be performed at each time step in these shock models
434: is that of the ionization balance and the corresponding mean cooling rate,
435: $\bar{\Lambda}$. The cooling and radiative emission is calculated using a very
436: large atomic data set which allows treatment of all ionization stages of
437: cosmically abundant elements up to fully ionized nickel. Using this atomic
438: data set, the rate equations for non-equilibrium ionization, recombination,
439: excitation, and radiative transfer and cooling are solved at each time step of
440: the flow. Details of the various collisional and radiative processes included
441: in MAPPINGS are described in \citet{sutherland1993}.
442:
443: Modeling of all fast radiative shocks is necessarily an iterative process
444: because the detailed structure of the shock depends on the ionization state of
445: the precursor gas entering the shock front. To calculate a fully
446: self-consistent model of a radiative shock a number of iterations are
447: required. Firstly an initial shock model is calculated using an estimate of
448: the ionization state of the precursor. Then the photoionized precursor is
449: calculated using the ionizing radiation field generated by the shock. This
450: process is then repeated, updating the ionization state of the precursor gas
451: at each iteration. The models presented here employ four such iterations,
452: which was found to be sufficient to allow the temperature and ionization state
453: of the precursor gas to stabilize at a constant value.
454:
455:
456:
457: In the final iteration of each model, the computation of the shock structure
458: is allowed to proceed until the gas has cooled to 1000K, below which no further significant
459: emission in the considered species is produced. The computation of
460: the precursor structure is terminated when the ionized fraction of hydrogen
461: falls to below 1 percent.
462:
463: Precursor components were computed for each individual model, in contrast to
464: DS96 who computed precursors only for each value of the shock velocity. The
465: precursor ionization does not generally depend on the magnetic parameter but
466: here we chose to compute each individual precursor to ensure full self
467: consistency between the shock and its precursor.
468:
469: \begin{figure*}
470: \centering
471: \includegraphics[scale=0.6]{f8a_col.ps} \\
472: \includegraphics[scale=0.6]{f8b_col.ps} \\
473: \includegraphics[scale=0.6]{f8c_col.ps} \\
474: \caption{The helium ionization structure, temperature profile and density
475: profile of models with solar abundance, equipartition magnetic field for shock
476: velocities of $v_{\rm s}$ = 200, 500, and 1000 km\ s$^{-1}$. The axes are the
477: same as for Figure~\ref{Hstruct_eg}. See the electronic edition of the
478: Journal for a color version of this figure.
479: \label{Hestruct_fig}}
480: \end{figure*}
481:
482:
483: \begin{figure*}
484: \centering
485: \includegraphics[scale=0.6]{f9a_col.ps} \\
486: \includegraphics[scale=0.6]{f9b_col.ps} \\
487: \includegraphics[scale=0.6]{f9c_col.ps} \\
488: \caption{The carbon ionization structure, temperature profile and density
489: profile of models with solar abundance, equipartition magnetic field for shock
490: velocities of $v_{\rm s}$ = 200, 500, and 1000 km\ s$^{-1}$. The axes are the
491: same as for Figure~\ref{Hstruct_eg}. See the electronic edition of the
492: Journal for a color version of this figure.
493: \label{Cstruct_fig}}
494: \end{figure*}
495:
496: \begin{figure*}
497: \centering
498: \includegraphics[scale=0.6]{f10a_col.ps} \\
499: \includegraphics[scale=0.6]{f10b_col.ps} \\
500: \includegraphics[scale=0.6]{f10c_col.ps} \\
501: \caption{The nitrogen ionization structure, temperature profile and density
502: profile of models with solar abundance, equipartition magnetic field for shock
503: velocities of $v_{\rm s}$ = 200, 500, and 1000 km\ s$^{-1}$. The axes are the
504: same as for Figure~\ref{Hstruct_eg}. See the electronic edition of the
505: Journal for a color version of this figure.
506: \label{Nstruct_fig} }
507: \end{figure*}
508:
509: \begin{figure*}
510: \centering
511: \includegraphics[scale=0.6]{f11a_col.ps} \\
512: \includegraphics[scale=0.6]{f11b_col.ps} \\
513: \includegraphics[scale=0.6]{f11c_col.ps} \\
514: \caption{The oxygen ionization structure, temperature profile and density
515: profile of models with solar abundance, equipartition magnetic field for shock
516: velocities of $v_{\rm s}$ = 200, 500, and 1000 km\ s$^{-1}$. The axes are the
517: same as for Figure~\ref{Hstruct_eg}. See the electronic edition of the
518: Journal for a color version of this figure.
519: \label{Ostruct_fig}}
520: \end{figure*}
521:
522: \begin{figure*}
523: \centering
524: \includegraphics[scale=0.6]{f12a_col.ps} \\
525: \includegraphics[scale=0.6]{f12b_col.ps} \\
526: \includegraphics[scale=0.6]{f12c_col.ps} \\
527: \caption{The sulfur ionization structure, temperature profile and density
528: profile of models with solar abundance, equipartition magnetic field for shock
529: velocities of $v_{\rm s}$ = 200, 500, and 1000 km\ s$^{-1}$. The axes are the
530: same as for Figure~\ref{Hstruct_eg}. See the electronic edition of the
531: Journal for a color version of this figure.
532: \label{Sstruct_fig} }
533: \end{figure*}
534:
535: \subsection{Model Grid Input Parameters}\label{sec:input}
536:
537: Each individual model in our shock library is defined by five physical
538: parameters; the pre-shock density, n, the shock velocity, $v_{s}$, the
539: pre-shock transverse magnetic field, B and the set of atomic abundances. The
540: library itself is organized into two main groups of model grids:
541:
542: \begin{itemize}
543: \item First, complete grids of models calculated for five different atomic
544: abundance sets; (depleted) Solar and 2$\times$Solar as used by
545: DS96 and which are based upon the older \citet{Anders89}
546: abundance set; a Solar abundance set based upon the \citet{Asplund05}
547: abundances as listed in \citet{dopita2005}, which is referred to
548: here as the `dopita2005' abundances; and an LMC and an
549: SMC abundance set as given by \citet{Russell92}. The abundances of the
550: individual elements (by number with respect to hydrogen) are listed for each
551: of these abundance sets in Table~\ref{abund_table}. Each of these model
552: grids assume a fixed pre-shock density of n=1
553: cm$^{-3}$, and consists of a set of models with shock velocities
554: covering the range 100 up to 1000 km
555: s$^{-1}$ in steps of 25 km s$^{-1}$, and magnetic fields of 10$^{-4}$, 0.5, 1.0, 2.0,
556: 3.23, 4.0, 5.0 and 10.0 $\mu$G. As the pre-shock density is unity, the
557: corresponding magnetic parameters are B/$\sqrt{\rm n}=10^{-4}$, 0.5, 1.0,
558: 2.0, 3.23, 4.0, 5.0 and 10.0 $\mu$G\ cm$^{3/2}$.
559:
560: \item Second, grids of models with solar abundance (as used by DS96),
561: calculated for densities of 0.01, 0.1, 1.0, 10, 100 and 1000 cm$^{-3}$. As
562: with the first set, each of these model grids covers shock velocities of 100
563: up to 1000 km\ s$^{-1}$ in steps of 25 km\ s$^{-1}$, and have the same
564: magnetic parameters of 10$^{-4}$, 0.5, 1.0, 2.0, 3.23, 4.0, 5.0 and 10.0
565: $\mu$G\ cm$^{3/2}$. However, as the magnetic field values required to obtain
566: these magnetic parameters are of course different for each density, additional
567: models were computed in order to be able to also
568: compare models of different
569: densities with the same transverse magnetic field. These are B$\sim10^{-3}$,
570: $\sim10^{-2}$, $\sim10^{-1}$, 1.0, 10 and 100 $\mu$G, calculated for
571: each density.
572:
573: \end{itemize}
574:
575: The magnetic field, B, and magnetic parameter, B/$\sqrt{\rm n}$, values are chosen
576: so as to cover the extremes expected in the ISM, while also sampling more
577: finely the magnetic field strengths which are near equipartition. Under
578: equipartition conditions the magnetic pressure is equal to the thermal
579: pressure, and the Alfv\'en speed is approximately equal to the gas sound
580: speed. Pressure equipartition occurs for ${\rm B}_{0}^{2}/4\pi \sim
581: n_0kT_0$ where
582: B$_{0}$ is the transverse magnetic field, and n$_0$, $T_0$ are the pre-shock
583: densities and temperatures \citep{dopita2003}. This condition is satisfied for
584: magnetic parameters B$/\sqrt{\rm n} \sim 3-5$. The value B$/\sqrt{\rm n} \sim 3.23$
585: was chosen as the nominal equipartition value.
586:
587: \begin{figure*}
588: \centering
589: \includegraphics[scale=0.6]{f13.ps} \\
590: \caption{Integrated hydrogen column densities for the shock and the precursor
591: structures. The column densities of the shock structure are shown for magnetic
592: parameters B/$\sqrt{\rm n}$=0.0001, 3.23 and 10. The curve labelled `PRECURSOR'
593: shows the column density of the precursor gas which does not depend on
594: the magnetic parameter. The dashed curves show the shock and precursor column
595: densities for the DS96 models, and the dotted curve shows the extrapolation of
596: the column density scaling relation for the shock column density of
597: DS96. \label{coldens_fig}}
598: \end{figure*}
599:
600: \begin{figure}
601: \centering
602: \includegraphics[scale=0.60]{f14.ps}
603: \caption{Total radiative fluxes of shocks as a function of shock velocity and
604: various values of the pre-shock densities.
605: The bold line is for the standard model having a solar abundance set, a
606: density n$=1$ cm$^{-3}$, and an equipartition pre-shock magnetic field. The
607: thinner lines near the bold line show the luminosities of the n$=1$ cm$^{-3}$,
608: equipartition models with abundances of twice solar, LMC and
609: SMC abundance sets. \label{lum_compare1_fig}}
610: \end{figure}
611:
612:
613: \begin{figure}
614: \centering
615: \includegraphics[scale=0.60]{f15.ps}
616: \caption{As for Figure \ref{lum_compare1_fig}, but showing the
617: precursor total radiative fluxes.
618: \label{lum_compare2_fig}}
619: \end{figure}
620:
621:
622: The input parameters for the all of the models are provided in
623: Table~\ref{param_table}. Each row of the table represents a velocity sequence
624: of models at a given abundance, density and magnetic field. The table is
625: organized into ten sets of models. The first five sets of models represent the
626: model grids for the five different abundances, and the following 5 sets of
627: models are the solar abundance models for different input densities and
628: magnetic fields. The number and range of model input parameters allows the
629: construction of various 2-D, or higher dimensional, grids of models.
630:
631:
632: \section{Ionizing Radiation Generated by the Shock}\label{sec:ionrad}
633:
634: The ionizing radiation produced in the cooling zone behind the shock front
635: shock is mostly composed of thermal bremsstrahlung (free-free) continuum and
636: resonance lines arising from many different elements and ionic stages. The
637: underlying exponential shape of the continuum is emphasized in
638: Figure~\ref{specfig1} where we show the ionizing spectra generated in the
639: n=1.0 cm$^{-3}$ solar abundance shock models. As can be seen,
640: higher velocity shocks result in
641: harder and more luminous ionizing spectra, with the spectral slope in
642: the log-linear plot clearly flattening with increasing velocity. Note
643: that the histogram nature of the figure also reveals the high energy
644: (frequency) resolution of the MAPPINGS spectral vector.
645: Figure \ref{specfig2} shows the ionizing spectra of
646: 400 and 1000 km\ s$^{-1}$ shocks on a $\nu F_{\nu}$ scale, illustrating both
647: the stronger UV fluxes and harder X-ray spectra generated in high velocity
648: shocks. This figure also shows a prominent low-temperature bound-free
649: continuum of hydrogen, produced in the cool, partially-ionized zone of the
650: recombination region of the shock, and the strong hydrogen two-photon
651: continuum produced mostly by the down-conversion of Ly$\alpha$ photons trapped
652: in this same region of the shock structure. Also present, though to a much
653: weaker scale, is the bound--free continuum arising from the heavier elements,
654: with the helium continuum the most obvious.
655:
656: \begin{figure*}[htb]
657: \centering
658: \includegraphics[scale=0.95]{f16a.ps}
659: \includegraphics[scale=0.95]{f16b.ps}
660: \caption{Line strength relative to H$\beta$ for 16 strong emission
661: lines between the Lyman Limit and 26$\mu$m as a function of shock
662: velocity. These are
663: computed for the shock component only, and are for the fiducial model with
664: solar abundance, n$=1.0$ cm$^{-3}$ and B$=3.23~ \mu$G. \label{M_n1_be_s_llf}}
665: \end{figure*}
666:
667:
668: \begin{figure*}[htb]
669: \centering
670: \includegraphics[scale=0.95]{f17a.ps}
671: \includegraphics[scale=0.95]{f17b.ps}
672: \caption{As Figure \ref{M_n1_be_s_llf} but for the precursor component
673: of the models only. \label{M_n1_be_p_llf} }
674: \end{figure*}
675:
676:
677: \begin{figure*}[htb]
678: \centering
679: \includegraphics[scale=0.95]{f18a.ps}
680: \includegraphics[scale=0.95]{f18b.ps}
681: \caption{As Figure \ref{M_n1_be_s_llf} but for the the full
682: shock+precursor model. \label{M_n1_be_sp_llf}}
683: \end{figure*}
684:
685: The strength of the ionizing field is a strong function of the shock velocity,
686: but does not significantly depend on either the atomic abundance or the
687: magnetic field. This is because the ionizing field is dominated by the
688: bremsstrahlung radiation, whose strength is controlled by the temperature and
689: density of the radiative zone, which is determined by the shock velocity and
690: pre-shock density. As magnetic field support in this cooling zone is
691: negligible, it has little affect on the emission, and, as hydrogen dominates
692: both electron and ion numbers for both the bremsstrahlung and free-bound
693: emission, changes in metallicity have only a small impact.
694:
695: \begin{figure*}
696: \centering
697: \includegraphics[scale=0.6]{f19a_col.ps}
698: \includegraphics[scale=0.6]{f19b_col.ps}
699: \caption{ The classic \cite{veilleux1987} [\ion{O}{3}]$\lambda$5007/H$\beta$
700: versus [\ion{N}{2}]$\lambda$6583/H$\alpha$ diagnostic diagram for the solar
701: abundance n$=1$ cm$^{-3}$ models. The left figure shows the shock-only
702: models, and the right figure shows the shock+precursor models. The grid is
703: comprised of lines of constant magnetic parameter shown with a bold linestyle,
704: and lines of constant shock velocity shown as the thin lines. The lines of
705: constant shock velocity are shown at 50 km~s$^{-1}$ intervals (and are color
706: coded with increasing red intensity for higher shock velocities in the
707: electronic edition). The velocity sequences are shown for all values of the
708: magnetic parameter (and are color coded with increasing blue intensity for
709: higher magnetic parameters in the electronic edition). Both the shock-only and
710: shock+precursor grids are shown for the shock velocities of 200--1000
711: km~s$^{-1}$, for which all the models have a fully ionized precursor. See the
712: electronic edition of the Journal for a color version of this figure.
713: \label{lrat1}}
714: \end{figure*}
715:
716: The hydrogen ionizing radiation flux, $L_{\rm UV}$, integrated for all
717: energies $h\nu > 13.6$\ eV, and over 2$\pi$\ sr is listed in
718: Table~\ref{ion_prop_table}. It is found to scale almost exactly as the
719: available enthalpy as:
720:
721: \begin{equation}
722: \label{Luv_eq}
723: {\rm L}_{\rm UV} = 2.44 \times 10^{-4} \left(\frac{v_S}{100\ {\rm km}\
724: {\rm s}^{-1}}\right)^{3.02} \times \left(\frac{\rm n}{{\rm
725: cm}^{-3}}\right) {\rm ergs}\ {\rm cm}^{-2}\ {\rm s}^{-1}
726: \end{equation}
727:
728: The ionizing fields may also be characterized in terms of their ionization
729: parameter in the pre-shock gas. As in DS96 we give ionization parameters,
730: $Q$, defined as the mean number of photons passing through unit area divided
731: by the total pre-shock particle density; $Q = N_{photons}/n_T$ (cm\
732: s$^{-1}$). Table~\ref{ion_prop_table} lists $Q$ in the \ion{H}{1}-ionizing
733: 13.6$< h\nu <$ 24 eV band, $Q({\rm H})$, the \ion{He}{1}-ionizing 24$< h\nu <$
734: 54 eV band, $Q({\rm He I})$, and the \ion{He}{2}-ionizing band $h\nu >$ 54 eV,
735: $Q({\rm HeII)}$ as a function of shock velocity. We also provide the
736: corresponding $R-$Type ionization front velocity $v_{ion}$\ and the
737: equilibrium electron temperatures, $T_e$ just ahead of the shock front.
738:
739: \section{Shock and Precursor Structures}\label{sec:struct}
740:
741: We now consider the ionization structures and the physical scales of
742: both the shock and precursor components. Figure \ref{Hstruct_eg} shows
743: the ionized and neutral hydrogen structures (as labelled) of the shock
744: and precursor components of the 500 km\ s$^{-1}$ solar abundance
745: model, along with the hydrogen density and electron temperature
746: profiles. Note that the hydrogen density is not plotted within the
747: precursor region as it is set at the constant value of n$_{\rm H}=1.0$
748: cm$^{-3}$ . The profiles are plotted in the frame of the shock front,
749: considered with respect to the time axis. This axis represents time
750: since (or prior to) the passage of the shock front, and can be
751: converted into a physical distance from the shock front via the shock
752: velocity.
753:
754:
755: In Figure~\ref{Hstruct_eg} (and subsequent ionization structure
756: Figures~\ref{Hstruct_V}--\ref{Sstruct_fig}), the structure of the precursor
757: is shown in the left panel and the shock structure is shown in the right
758: panel. The shock front is located at time $t=0$. The time taken to reach
759: equipartition in the electrons, ions, and un-ionized gas is assumed to be
760: negligible, so that the shock is therefore unresolved and the density and
761: electron temperature jump discontinuously at the shock front. Note also that
762: the time axes are shown in units of 10$^{12}$s, but the scales and ranges are
763: necessarily different for the shock and precursor panels of these diagrams, as
764: well as between the diagrams with different parameters.
765:
766: As described in DS96 the a number of zones can be identified in the shock
767: structures. These are indicated in Figure~\ref{Hstruct_eg} and are,
768: respectively:
769:
770: \begin{enumerate}
771: \item \emph{The ionization region.} This region, just after the shock
772: front, is in an ionization state below that appropriate for the
773: electron temperature assuming collisional ionization equilibrium
774: (CIE), due to discontinuous temperature jump at the front. After
775: the passing of the front, the gas rapidly (barely resolved in the
776: figures) adjusts from the pre-shock ionization state to the
777: appropriate CIE state for the post-shock temperature. The rate at
778: which it adjusts depends upon the collisional ionization rates of
779: these species. As discussed in DS96, this region has strong line
780: emission due to this state of high temperature and relatively
781: low-ionization.
782:
783: \item \emph{The high temperature radiative zone.} This is the zone in which
784: most of the EUV and soft X-ray flux is emitted. The ionization state
785: is approximately in coronal equilibrium for the temperature of the
786: region. Even though the cooling rate of this region is low due to
787: the high temperature, it is
788: the dominant contributor to the radiation field of the shock,
789: assisted by the high-ionization, optically-thin state of the gas.
790:
791: \item \emph{The non-equilibrium cooling zone.} Once the temperature of
792: the post shock gas drops below $\sim 10^6$ K, the cooling rate
793: becomes very high and the recombination timescales of a large number
794: of ions becomes longer than the local cooling timescale. As a
795: result, the ionization state lags and the plasma is in a higher
796: degree of ionization than collisional ionization equilibrium would
797: suggest. The relatively high ionization state for the temperature
798: means that the collisional
799: line emission is weak for these intermediate ionization
800: species, and that the gas remains optically thin to the diffuse
801: radiation of the previous region.
802:
803: \item \emph{The super-cooled zone} is the region where photoionization
804: starts to become important in determining the ionization balance.
805: Initially, however, the ionization state is still too high to
806: efficiently absorb the ionizing photons. This leads to an
807: over-cooling, and the temperature falls below the value given by
808: photoionization equilibrium. The width of this region is dependent
809: upon both the recombination and photoionization timescales of the ions.
810:
811: \item \emph{The photoabsorption and recombination zone.} This is essentially
812: an equilibrium plane-parallel \ion{H}{2} region illuminated by the
813: downstream EUV photon field. The density in this region is much higher than
814: the pre-ionized region ahead of the shock, so that the effective ionization
815: parameter in this zone is much lower, and the \ion{H}{2} region is very much
816: thinner. Eventually the EUV photons are absorbed, and hydrogen finally
817: recombines. However, there is a cool, partially ionized region behind the
818: main recombination front. This has a temperature of a few thousand degrees,
819: and is ionized by the hard penetrating X-rays which undergo Auger processes
820: and these lead to the emission of fast electrons which in turn produce
821: secondary ionization cascades.
822:
823: \end{enumerate}
824:
825: \begin{figure}[htb]
826: \includegraphics[scale=0.6]{f20.ps}
827: \caption{ Comparison of the new \mapiii\ models with the DS96 on the
828: [\ion{O}{3}]$\lambda$5007/H$\beta$ versus [\ion{N}{2}]$\lambda$6583/H$\alpha$
829: diagram. The DS96 models are shown in black, they are labelled with the shock
830: velocity in km~s$^{-1}$, and include velocity sequences for magnetic
831: parameters of 0, 1, 2 and 4 $\mu$G\ cm$^{3/2}$ which increase from left to right
832: in both the shock-only (lower) and shock+precursor (upper) grids. The new
833: \mapiii\ models are shown in greyscale and cover the same range in shock
834: velocity and magnetic field as shown in Figure~\ref{lrat1}. Note that the new
835: diagrams show systematic changes with respect to the earlier ones, and that
836: the new models show that these ratios become insensitive to shock velocity for
837: $v_{\rm s} > 500$ km~s$^{-1}$.
838: \label{lrat4}}
839: \end{figure}
840:
841:
842: \begin{figure*}[htb]
843: \includegraphics[scale=0.6]{f21a_col.ps}
844: \includegraphics[scale=0.6]{f21b_col.ps}
845: \caption{ Shock and shock+precursor model grids for the five different atomic
846: abundance sets used in the shock model library for the [\ion{O}{3}]/H$\beta$
847: versus [\ion{N}{2}]/H$\alpha$ diagram. The shock-only models are shown in (a),
848: and the shock+precursor models are shown in (b). All model grids are shown for
849: $v_{\rm s}$ 200--1000 km~s$^{-1}$, with lines of constant shock velocity drawn
850: at intervals of 50 km~s$^{-1}$. Each grid is labelled with the abundance set
851: used, and generally moves from left to right as the total metallicity
852: increases. See the electronic edition of the Journal for a color version of
853: this figure.
854: \label{lrat2}}
855: \end{figure*}
856:
857: \subsection{Hydrogen ionization structures}
858:
859: Figures \ref{Hstruct_V} - \ref{Hstruct_abund} show, respectively, the
860: effects of changing velocity, magnetic field, pre-shock density and abundance
861: on the structures of the hydrogen ionization fraction, hydrogen number
862: density, n$_{\rm H}$, and the electron temperature, $T_{\rm e}$, in the shock
863: and precursor regions. Note that in each diagram the time scale changes due to
864: the effects of the changing parameters. In each diagram the parameters
865: of the model (abundance set, velocity, density, and magnetic field
866: respectively) are listed in the top right corner for clarity.
867:
868:
869:
870: \begin{deluxetable*}{lrrrrrr}
871: \tablecolumns{7}
872: \tablewidth{0pc}
873: \tablecaption{Shock Ionizing Properties \label{ion_prop_table}}
874: \tablehead{
875: \colhead{v$_{s}$} &
876: \colhead{${\rm L}_{\rm UV}$} &
877: \colhead{Q(H \sc{I})} &
878: \colhead{Q(He \sc{I})} &
879: \colhead{Q(H \sc{II})} &
880: \colhead{v$_{\rm ion}$} &
881: \colhead{T$_e$} \\
882: \colhead{(km\ s$^{-1}$)} &
883: \colhead{(ergs\ cm$^{-2}$s$^{-1}$)} &
884: \colhead{(km\ s$^{-1}$}) &
885: \colhead{(km\ s$^{-1}$}) &
886: \colhead{(km\ s$^{-1}$}) &
887: \colhead{(km\ s$^{-1}$}) &
888: \colhead{(K)} }
889: \startdata
890: 200 & 8.041E-4 & 209.1 & 43.0 & 4.9 & 214.2 & 7033 \\
891: 300 & 3.519E-3 & 430.8 & 312.1 & 83.1 & 688.3 & 10288 \\
892: 400 & 9.106E-3 & 664.3 & 598.7 & 513.2 & 1480.0 & 12145 \\
893: 500 & 1.751E-2 & 985.2 & 1195.0 & 602.2 & 2318.3 & 15435 \\
894: 600 & 2.968E-2 & 1158.2 & 1628.2 & 800.7 & 2989.2 & 18486 \\
895: 700 & 4.631E-2 & 1559.2 & 1986.0 & 1094.0 & 3865.8 & 20516 \\
896: 800 & 6.815E-2 & 1958.8 & 2362.3 & 1542.7 & 4885.8 & 21934 \\
897: 900 & 9.645E-2 & 2375.0 & 2920.4 & 2206.6 & 6251.7 & 23389 \\
898: 1000 & 1.318E-1 & 2740.6 & 3262.7 & 2786.8 & 7325.0 & 24760
899: \enddata
900: \end{deluxetable*}
901:
902: The dominant parameter of the shock (and precursor) structure and emission is
903: the velocity, as is seen in Figure~\ref{Hstruct_V}. This figure shows our
904: fiducial n$=1.0$ cm$^{-3}$, B$/\sqrt{\rm n}=3.23~\mu$G cm$^{3/2}$, solar
905: abundance model with two velocities; $v_{\rm s}=200$ and 1000 km\,s$^{-1}$
906: (three including the $v_{\rm s}=500$ km\,s$^{-1}$ shown in Figure
907: \ref{Hstruct_eg}). There is a clear increase of both the extent and
908: temperature of the radiative zone as the velocity increases due to the greater
909: shock strength, leading to the changes in the ionizing radiation field shown in
910: Figures \ref{specfig1} and \ref{specfig2}. This increase of both the hardness
911: and luminosity of the radiation field leads to the greater extent and
912: ionization in the precursor region, visible through the increased temperature
913: in this figure and the higher ionization species seen in Figures
914: \ref{Hestruct_fig}--\ref{Sstruct_fig}. The increased radiation field also
915: affects the ionization of the post-shock recombination region. This is
916: mollified somewhat however by the density increase associated with the faster
917: shock, which leads to a smaller increase in the ionization parameter, though
918: the effects of the harder radiation are still visible.
919:
920:
921: \begin{figure*}
922: \includegraphics[scale=0.6]{f22a_col.ps}
923: \includegraphics[scale=0.6]{f22b_col.ps} \\
924: \includegraphics[scale=0.6]{f22c_col.ps}
925: \includegraphics[scale=0.6]{f22d_col.ps}
926: \caption{Shock and shock+precursor model grids of solar abundance models for
927: two different pre-shock densities on the [\ion{O}{3}]$\lambda$5007/H$\beta$
928: versus [\ion{N}{2}]$\lambda$6583/H$\alpha$ diagram. The top figures show the
929: high density case and the lower figures the low density case, with a) The
930: n$=1000$ cm$^{-3}$ shock model grid, b) the n$=1000$ cm$^{-3}$ shock+precursor
931: models, c) The n$=0.01$ cm$^{-3}$ shock model grid, and d) the n$=0.01$
932: cm$^{-3}$ shock+precursor models. The grids show shock velocities over the
933: range $v_{\rm s}$=200-1000 km~s$^{-1}$ with lines of constant magnetic
934: parameter shown as thick lines (and colored with increasing blue intensity for
935: higher magnetic parameter in the electronic edition). The thin lines represent
936: constant shock velocity and these are plotted for the full range of magnetic
937: parameters for each density as listed in Table~2 (and are shown colored with
938: increasing red intensity for higher shock velocities in the electronic
939: edition). See the electronic edition of the Journal for a color version of
940: this figure.
941: \label{lrat3}}
942: \end{figure*}
943:
944:
945: Figure~\ref{Hstruct_B} shows models with n=1.0 cm$^{-3}$, $v_{\rm s}=500$
946: km\,s$^{-1}$, and solar abundance for differing magnetic fields. The top panel
947: shows a model with the magnetic parameter set to B$/\sqrt{\rm n}=0.0001 \mu$G
948: cm$^{3/2}$, and the bottom panel has B$/\sqrt{\rm n}=10.0 \mu$G cm$^{3/2}$,
949: which can be compared with the equipartition model with B$/\sqrt{\rm n}=3.23
950: \mu$G cm$^{3/2}$ model in Figure \ref{Hstruct_eg}. In all three cases the
951: precursor structure is almost unchanged, as the radiation field and luminosity
952: arising from the shock are little affected by changes in B (except at
953: extremely high values), as discussed in sections \S \ref{sec:ionrad} and \S
954: \ref{sec:lum}. The insensitivity of the radiation field to B can be seen in
955: this figure through the similarity of the radiative zone between the
956: models. This similarity, a result of the insensitivity of the temperature jump
957: to the magnetic field, means that the ionizing bremsstrahlung radiation that
958: dominates the energetics is practically the same. However, what is sensitive
959: to the transverse magnetic field is the density in the recombination zone
960: behind the shock front. As discussed in DS96, the maximum compression in this
961: post-shock region is driven by the ratio of the post- to pre-shock magnetic
962: fields. Thus a higher initial magnetic field results in a lower density
963: post-shock gas, and therefore, given the similarity in radiation field, a
964: higher effective post-shock ionization parameter. This results in a
965: significantly different ionization and emission structure in the post-shock
966: gas, leading to the strong diagnostics seen in DS95 and in this work (\S
967: \ref{sec:specsig}).
968:
969: Figure \ref{Hstruct_dens} shows the effect of changing the pre-shock
970: density. We plot our fiducial solar abundance models, with $v_{\rm s}=500$
971: km\,s$^{-1}$ and B$/\sqrt{\rm n}=3.23~\mu$G cm$^{3/2}$, for each of the
972: densities n=0.1, 10,
973: and 1000 cm$^{-3}$, from top to bottom (with the n=1.0 cm$^{-3}$ case shown in
974: Figure \ref{Hstruct_eg}). Note that the magnetic field, B, is also varied in
975: order to keep the magnetic parameter constant for the models shown.
976: This figure demonstrates that at a given shock
977: velocity, the region influenced by the shock scales with density (to the
978: extent that collision de-excitation effects are unimportant), proving
979: explicitly that to first order the product n$_0t$ is an invariant with
980: changing density, as asserted by DS96.
981:
982: \begin{figure*}[htb]
983: \includegraphics[scale=0.6]{f23a_col.ps}
984: \includegraphics[scale=0.6]{f23b_col.ps}
985: \caption{ The \cite{veilleux1987} plot of [\ion{O}{3}]$\lambda$5007/H$\beta$
986: versus [\ion{S}{2}]$\lambda\lambda$6716,31/H$\alpha$ diagram for a) the shock
987: and b) the shock + precursor models with n$=1$\ cm$^{-3}$ and solar abundance.
988: The range and step size of the shock velocity and magnetic parameter for these
989: grids are the same as in Figure~\ref{lrat1}. See the electronic edition of
990: the Journal for a color version of this figure.
991: \label{lrat5}}
992: \end{figure*}
993:
994: \begin{figure*}[htb]
995: \includegraphics[scale=0.6]{f24a_col.ps}
996: \includegraphics[scale=0.6]{f24b_col.ps}
997: \caption{The temperature sensitive [\ion{O}{3}]$\lambda$4363/
998: [\ion{O}{3}]$\lambda$5007 ratio ($R_{\rm OIII}$) versus the
999: [\ion{O}{3}]$\lambda$5007/H$\beta$ line ratio diagram for the solar abundance
1000: n$=1$\ cm$^{-3}$ models. Panel (a) shows the shock grid, and panel (b) shows
1001: the shock+precursor grid. The range and step size of the shock velocity and
1002: magnetic parameter for these grids are the same as in Figure~\ref{lrat1}. See
1003: the electronic edition of the Journal for a color version of this figure.
1004: \label{lrat6}}
1005: \end{figure*}
1006:
1007: Changes in the atomic abundances directly affect the cooling rate.
1008: Figure~\ref{Hstruct_abund} shows the equipartition case, B/$\sqrt{\rm n}$ =
1009: 3.23~$\mu$G cm$^{3/2}$, with n=1.0 cm$^{-3}$ and $v_{\rm s}=500$
1010: km\,s$^{-1}$ for models
1011: using the 2$\times$Solar, dopita2005, and SMC abundances. All three models
1012: differ in structure, with the expected clear progression in cooling efficiency
1013: as we increase in metallicity from the SMC abundances to the twice Solar. This
1014: difference in efficiency leads to the changes visible in the spatial and
1015: temporal extent of the models, in both the precursor and shock regions. The
1016: changes in metallicity also affect the ionizing radiation from the shock due
1017: to the weakening of the heavy element free-bound features. However, as
1018: discussed in section \ref{sec:ionrad}, the effect is minimal.
1019:
1020: \subsection{Ionization Structures of Abundant Elements}
1021:
1022: The shock model library includes the full fractional ionization structure of
1023: all elements listed in Table~\ref{abund_table}. In Figures
1024: \ref{Hestruct_fig}--\ref{Sstruct_fig} we show the ionization structures of
1025: helium, carbon, nitrogen, oxygen and sulfur for three velocities (200, 500 and
1026: 1000 km\ s$^{-1}$) of the solar abundance, n$=1.0$ cm$^{-3}$ models with the
1027: equipartition magnetic parameter. The associated hydrogen structures are shown
1028: in Figures \ref{Hstruct_eg} and \ref{Hstruct_V}. In each figure, the
1029: ionization fraction, $X_{\rm ion}$, is shown at the bottom (with values $\le
1030: 1$), with each ionization state labelled and the odd and even ionization
1031: states shown as the solid and dot-dashed curves respectively for clarity. Also
1032: included in each diagram is the hydrogen density, n$_{\rm H}$, and electron
1033: temperature structure, T$_{\rm e}$, to trace the different shock and precursor
1034: zones.
1035:
1036:
1037: Together, these figures elucidate further what has been discussed in the
1038: previous sections; faster shocks lead to a higher temperature, and therefore
1039: higher ionization, radiative zone. This in turn leads to a more luminous and
1040: harder ionizing spectrum as can be traced within the precursor zone. Similarly
1041: this figure allows one to trace the more complex ionization structure of
1042: the post-shock recombination region, affected both by the increased ionizing
1043: spectrum and increased density that results from faster shocks.
1044:
1045: \begin{figure*}[h]
1046: \centering
1047: \includegraphics[scale=0.55]{f25a_col.ps}
1048: \includegraphics[scale=0.55]{f25b_col.ps}
1049: \caption{The ultraviolet carbon line ratio diagnostic plot for the solar
1050: abundance n$=1$\ cm$^{-3}$ shock (a) and shock+precursor models (b). The
1051: shock only grid enables both the shock velocity and the magnetic parameter to
1052: be independently determined, but the shock + precursor grid is multi-valued at
1053: many positions. The range and step size of the shock velocity and magnetic
1054: parameter for these grids are the same as in Figure~\ref{lrat1}. See the
1055: electronic edition of the Journal for a color version of this figure.
1056: \label{lrat7}}
1057: \end{figure*}
1058:
1059: However the true strength of these figures, and the library in general, is
1060: their diagnostic ability, allowing the reader to trace exactly where each
1061: ionization state arises, which is especially interesting given the X-ray
1062: ionized regions that occur far from the ionization front in both the precursor
1063: and post-shock regions.
1064:
1065:
1066: \subsection{Column Densities}
1067:
1068: The column densities of the shock and precursor components of the models
1069: inevitably increase with shock velocity, and can be as large as
1070: $N$(H)=4$\times10^{21}$~cm$^{-2}$ in the shock itself for the highest velocity
1071: solar models. The precursor column densities range from between 3 and 20 times
1072: larger than the corresponding shock component, with this ratio decreasing for
1073: the higher velocity shocks.
1074:
1075:
1076: Figure \ref{coldens_fig} displays the integrated \ion{H}{2} column density of
1077: the shock as a function of shock velocity. This is shown for the solar
1078: abundance n$=1.0$ cm$^{-3}$ models with equipartition magnetic parameter, as
1079: well as the for maximum and minimum magnetic parameters of 0.0001 and
1080: 10.0. The dashed line shows the scaling relation given by DS96, and the dotted
1081: line shows the extrapolation of this beyond shock velocities of 500
1082: km~s$^{-1}$. The diagram shows a weak dependence of the column density on the
1083: magnetic parameter, and a decrease in the slope beyond shock velocities of 700
1084: km~s$^{-1}$.
1085:
1086: The column density of the precursor does not depend on the magnetic
1087: parameter. All solar abundance, n=1.0 cm$^{-3}$ models fall on the same
1088: curve, marked `PRECURSOR' in Figure~\ref{coldens_fig}. There is however a
1089: systematic difference in the precursor column density compared to the DS96
1090: models, with the newer models having a $\sim$10\% decrease in column density,
1091: consistent with the slightly lower fraction of ionizing flux compared with
1092: total flux produced in the cooling plasma in the newer models.
1093:
1094: Tables of the column densities for all species for every library model are
1095: available on-line (see section~\ref{online}). These columns are a function of
1096: both the hydrogen column density and the ionization structures seen in the
1097: previous figures. Example tables giving the column densities of the shock and
1098: precursor components of the n$=1.0$ cm$^{-3}$ solar abundance models at
1099: velocities of 200, 500 and 1000 km~s$^{-1}$ models are presented in
1100: Tables~\ref{coldens_tab3} and 5.
1101:
1102: \begin{figure*}[h]
1103: \centering
1104: \includegraphics[scale=0.6]{f26.ps}
1105: \caption{ The \ion{C}{3}]$\lambda$1909/\ion{C}{2}]$\lambda2326$ with
1106: \ion{Ne}{3}$\lambda$3869/\ion{Ne}{5}]$\lambda$3426 diagram. The shock model
1107: grid is shown in light gray (cyan in the electronic edition) and is labelled
1108: with the shock velocity. The shock+precursor grid is shown in dark grey (green
1109: in the electronic edition), and displays a twisted shape on these axes, and
1110: also overlaps the shock grid. The dusty and dust-free models of
1111: \cite{groves2004b} are also shown with the ionization parameter increasing to
1112: the left of the diagram before the turn-over in these curves. The A$_{\rm
1113: M/I}$ models from \cite{bws96} are shown as a dotted line with the + symbols
1114: indicating increments of 0.25 dex. The data points are those from
1115: \cite{best2000} and \cite{Inskip02} as indicated in the text.
1116: \label{UV_C_Ne_diag}}
1117: \end{figure*}
1118:
1119: \begin{figure*}
1120: \includegraphics[scale=0.6]{f27a_col.ps}
1121: \includegraphics[scale=0.6]{f27b_col.ps}
1122: \caption{ IR diagnostic diagram of
1123: [\ion{Ne}{5}]14.3$\mu$m/[\ion{Ne}{2}]12.8$\mu$m versus
1124: [\ion{Ne}{3}]15.5$\mu$m/[\ion{Ne}{2}]12.8$\mu$m. Panels (a) and (b) show
1125: respectively the shock and shock+precursor model grids for the solar abundance
1126: n$=1$\ cm$^{-3}$ models. The range and step size of the shock velocity and
1127: magnetic parameter for these grids are the same as in Figure~\ref{lrat1}. See
1128: the electronic edition of the Journal for a color version of this figure.
1129: \label{Ne_IR_diag}}
1130: \end{figure*}
1131:
1132:
1133: \subsection{Radiative fluxes of Shock and Precursor Components}\label{sec:lum}
1134:
1135: A fully-radiative shock will, by definition, radiatively dissipate all of the
1136: energy flux through the shock. Thus the total radiative flux, $L_{\rm Tot}$,
1137: is equal to $0.5 \rho v_{s}^3$. DS96 provided scaling relations for the total
1138: luminosity of a shock, and showed a similar scaling exists for the total
1139: luminosity of the precursor emission. This follows because, in the precursor,
1140: all of the radiation emitted in the pre-shock direction is eventually
1141: processed.
1142:
1143: \begin{figure*}[htb]
1144: \includegraphics[scale=0.6]{f28a_col.ps}
1145: \includegraphics[scale=0.6]{f28b_col.ps}
1146: \caption{ IR diagnostic diagram of
1147: [\ion{Ne}{5}]14.3$\mu$m/[\ion{Ne}{2}]12.8$\mu$m versus
1148: [\ion{O}{4}]25.9$\mu$m/[\ion{Ne}{2}]12.8$\mu$m. Panels (a) and (b) show
1149: respectively the shock and shock+precursor model grids for the solar abundance
1150: n$=1$\ cm$^{-3}$ models. The range and step size of the shock velocity and
1151: magnetic parameter for these grids are the same as in Figure~\ref{lrat1}. See
1152: the electronic edition of the Journal for a color version of this figure.
1153: \label{Ne_O_IR_diag}}
1154: \end{figure*}
1155:
1156:
1157: In Figures~\ref{lum_compare1_fig} and \ref{lum_compare2_fig} we
1158: show the total radiative fluxes generated in the shock and in the precursor
1159: respectively, in the \mapiii\ models with solar abundance and
1160: equipartition magnetic fields. Here we cover the full range of pre-shock
1161: density ($0.01-1000$ cm$^{-3}$). To illustrate the relatively minor
1162: adjustments to the total radiated flux due to different abundances we also
1163: show the luminosities for the models with 2$\times$Solar, Dopita2005, SMC and
1164: LMC abundance sets for n$=1.0$ cm$^{-3}$. Changing the magnetic field from the
1165: equipartition value produces too small an effect on the total radiative flux
1166: to be visible on the scale of these plots.
1167:
1168: \section{Spectral Signatures of the Shock and Precursor Gas}\label{sec:specsig}
1169:
1170: While the total luminosities of the shock effectively depends only upon the
1171: density and the shock velocity, the detailed emission line spectra depend
1172: strongly on the physical and ionization structure of the shock. This is
1173: determined primarily by the shock velocity and the magnetic
1174: parameter. However, at the higher pre-shock densities (and at lower values of
1175: the magnetic parameter) the density close to the photoionized tail and the
1176: recombination zone of the shock becomes sufficiently high for collisional
1177: de-excitation of forbidden lines to become important. In these circumstances,
1178: the emission line spectrum of the shock becomes dependent upon the density as
1179: well.
1180:
1181: \begin{figure}[h]
1182: \includegraphics[scale=0.6]{f29_col.ps}
1183: \caption{ IR diagram of [\ion{S}{4}]10.44$\mu$m/[\ion{S}{3}]18.7$\mu$m versus
1184: [\ion{Ne}{3}]15.5$\mu$m/[\ion{Ne}{2}]12.8$\mu$m showing the shock and shock+precursor
1185: grids of the solar abundance n$=1$\ cm$^{-3}$ models, as labelled. The range
1186: and step size of the shock velocity and magnetic parameter for these grids are
1187: the same as in Figure~\ref{lrat1}. See the electronic edition of the Journal
1188: for a color version of this figure. \label{S_Ne}}
1189: \end{figure}
1190:
1191: Shocks are characterized by regions of high electron temperature and
1192: ionization state. As a consequence, they display a rich spectrum of
1193: collisionally excited UV lines. The shock velocity controls the shape of the
1194: ionizing spectrum produced by the shock, and the magnetic parameter controls
1195: the effective ionization parameter in the photoionized tail of the
1196: shock. Higher ionization parameters yield a higher mean ionization in the
1197: plasma, so therefore faster shocks give an (optical) spectrum of somewhat
1198: higher ionization.
1199:
1200: The emission line ratios (with respect to H$\beta$=1) of a selection of lines
1201: for the shock, precursor and shock+precursor components of the solar
1202: abundance, n=1.0 cm$^{-3}$ models are shown separately in
1203: Figures~\ref{M_n1_be_s_llf}-\ref{M_n1_be_sp_llf} as a function of shock
1204: velocity.
1205:
1206:
1207:
1208:
1209: As discussed before, the precursor emission can be considered as an \ion{H}{2}
1210: region with high ionization parameter. In this region strong cooling
1211: lines, like [\ion{O}{3}]$\lambda$5007, are
1212: strong and generally increase in strength with shock velocity
1213: as the spectrum produced by the cooling plasma becomes harder, and its flux
1214: increases. Eventually, as the shock velocity is increased, the average state
1215: of ionization in the precursor region becomes high and the relative
1216: intensity of the lines produced by lower ionization turn over,
1217: flattening and even decreasing at the higher velocities. For example,
1218: in Figure \ref{M_n1_be_p_llf} [\ion{O}{3}]$\lambda$5007 can be seen to
1219: flatten out around 600 km~s$^{-1}$, and
1220: even decrease at 1000 km~s$^{-1}$.
1221:
1222: In addition the emission line ratios of 40 strong lines are tabulated for
1223: shock, precursor and shock+precursor in Tables 6, 7, and 8 respectively. Complete tables of
1224: emission line ratios of all lines, for all models are available electronically
1225: ( \emph{see} section~\ref{online}).
1226:
1227: \subsection{Line Ratio Diagrams}
1228:
1229: Diagnostic emission line ratio diagrams are well established as a powerful and
1230: practical way to investigate the physics of emission line gas. At a basic
1231: level they can be used empirically to identify the key excitation mechanism, and
1232: divide active galaxies from their star-formation dominated counterparts. At
1233: a deeper level, grids of theoretical models can be used to determine chemical
1234: abundances and derive physical parameters such as electron temperature,
1235: density, ionization parameter or shock velocity. Various sets of diagnostic
1236: diagrams have been proposed and used - \cite{baldwin1981} (optical),
1237: \cite{veilleux1987} (optical-IR), \cite{allen1998} (near-UV) and
1238: \cite{groves2004b} (UV, optical, IR). These 2-D diagrams remain useful
1239: alongside multi-dimensional methods for comparing data to models because the
1240: ratios can be chosen to be independent of reddening, and tuned to be sensitive
1241: to particular physical properties of interest. For example, \cite{Dopita06}
1242: have provided theoretical strong emission line diagnostics for ensembles of
1243: \ion{H}{2} regions to enable more accurate measurement of chemical abundances
1244: in unresolved starburst galaxies, and \cite{kewley2006} combined standard line
1245: ratio diagrams with velocity dispersion measurement to help refine the
1246: classification of active galaxies.
1247:
1248:
1249: As part of the shock model library available online, we include interactive
1250: plotting tools to plot diagnostic diagrams, such as the examples
1251: presented in the following sections. These tools enable a
1252: much clearer picture of these diagnostics than possible within the
1253: paper, especially with those grids that
1254: twist and fold upon themselves in the 2-D diagrams.
1255: In addition, many other diagnostic ratios than those presented here
1256: can be formed to gain insights into what diagnostics can be used when
1257: only a limited sample of lines are available.
1258:
1259: \subsection{Optical Diagnostics}
1260:
1261:
1262: As a first example, we plot the new shock model grids on a number of the
1263: standard optical diagnostic diagrams to reveal the general shape of the model
1264: grids, and the sensitivity of the diagnostics to the parameters of shock
1265: velocity, magnetic parameter and chemical abundance set. Figure~\ref{lrat1}
1266: shows the perhaps the most frequently used of the diagnostic diagrams, the
1267: \cite{veilleux1987} plot of [\ion{O}{3}]$\lambda$5007/H$\beta$ versus
1268: [\ion{N}{2}]$\lambda$6583/H$\alpha$, also commonly known as the BPT
1269: diagram \citep{baldwin1981}. For clarity, the shock and shock+precursor grids for
1270: the solar abundance n$=1$ cm$^{-3}$ models are plotted separately but
1271: on the same scale, as is the case in most of the following diagrams,
1272: due to the overlap of the two grids. These two
1273: grids represent the physical extremes of having little or no gas ahead of the
1274: shock (shock-only), and having an extensive, radiation bounded, precursor
1275: region ahead of the shock. As in DS95, the two grids show the range in
1276: shock velocity and magnetic parameter, as labelled in the diagram. The
1277: lines of constant velocity are plotted at 50 km\ s$^{-1}$ intervals as
1278: the thin lines.
1279: The lines of constant magnetic parameter are plotted as the
1280: thick lines in the grid, with the value of the magnetic parameter as
1281: marked. For a clearer picture of this and all other diagnostic
1282: diagrams in the paper we provide color figures in the online journal.
1283:
1284: \begin{figure}
1285: \includegraphics[scale=0.6]{f30.ps}
1286: \caption{ IR diagram [\ion{Fe}{2}]26.0$\mu$m/[\ion{O}{4}]25.9$\mu$m versus
1287: [\ion{O}{4}]25.9$\mu$m/[\ion{Ne}{2}]12.8$\mu$m as introduced by
1288: \cite{lutz2003}, overlaid with the shock and shock+precursor model
1289: grids. \label{lutz_fig}}
1290: \end{figure}
1291:
1292:
1293: This figure reveals a similar picture to that in DS95 (their Fig.~2b),
1294: with the shock-only and shock+precursor grids occupying different
1295: regions of the diagram, a result of the different processes discussed
1296: in the previous sections. Both grids shows broad steps in velocity
1297: up to a velocity of $\sim550$\kms, at which the grids begin to
1298: compress and turn upon themselves, becoming degenerate. However, for
1299: the shock-only grid an increase in velocity causes an increase in the
1300: [\ion{N}{2}]/H$\alpha$ ratio, while for the shock+precursor models in
1301: results in an increase of the [\ion{O}{3}]/H$\beta$ ratio. The
1302: situation is reversed for an increase of the magnetic parameter,
1303: causing the spread observed in the grids, though the shock+precursor
1304: models tend to be degenerate in this parameter at high values. The
1305: spread of the grids and their separation indicate the diagnostic power
1306: of these figures, allowing the determination of the physical
1307: parameters in known shock-ionized plasmas.
1308:
1309: \begin{figure}[h]
1310: \includegraphics[scale=0.6]{f31.ps}
1311: \caption{Comparison of the shock and shock+precursor models to SDSS line
1312: ratios on the [\ion{O}{3}]$\lambda$5007/H$\beta$ versus
1313: [\ion{N}{2}]$\lambda$6583/H$\alpha$ diagram. The model grids are those with
1314: 2$\times$solar abundance. The dotted and dashed curved lines represent the
1315: \cite{kewley2001} and \cite{kauffmann2003} classification lines respectively
1316: which divide \ion{H}{2} regions (lower left) from the region occupied by AGN.
1317: \label{sdss_compare1}}
1318: \end{figure}
1319:
1320: Improvements to the \mapiii\ code compared to the \mapii\ code used by DS96
1321: means that there are some significant and systematic differences compared to
1322: the DS95 grids. Figure \ref{lrat4} shows a direct comparison of the DS95 shock
1323: and shock+precursor model grids overlaid on our new solar abundance shock and
1324: shock+precursor models for the same diagnostic diagram. The improvement of
1325: our new models in terms of parameter space is clear, with DS95,96 models
1326: limited to 500 km\ s$^{-1}$ in velocity space, and only covering magnetic
1327: parameters of 0--4 $\mu$G cm$^{3/2}$. The shock component of the models agree
1328: reasonably well over the common parameter ranges, although appears somewhat
1329: offset. However the shock+precursor grid shows a difference of up to 0.5 dex
1330: in the [\ion{O}{3}]$\lambda$5007/H$\beta$ and
1331: [\ion{N}{2}]$\lambda$6583/H$\alpha$ ratios. We note again that on this diagram
1332: the new shock+precursor models turn over at high shock velocity. One
1333: consequence of this is that some active galaxies such as high-redshift radio
1334: galaxies, which have been confirmed as being excited by shocks
1335: and which were interpreted as matching 400-500 km\ s$^{-1}$
1336: shock+precursor models, may in fact be characterized by higher shock
1337: velocities. This would help resolve the mis-match between radial velocity
1338: dispersions and shock velocities inferred by line ratios as inferred, for
1339: example, by \citet{Reuland07}.
1340:
1341: Figure \ref{lrat2} shows the same diagram for the different chemical
1342: abundance sets which we have used. Each grid
1343: covers the same range in shock velocity and magnetic parameter. In
1344: both shock and shock+precursor models the
1345: [\ion{O}{3}]$\lambda$5007/H$\beta$ ratio is only weakly affected by
1346: abundance, as the increased temperatures within the models from the
1347: reduced cooling counteracts somewhat the effects of the drop in oxygen
1348: abundance on this ratio (see Figure \ref{Hstruct_abund}). The
1349: [\ion{N}{2}]$\lambda$6583/H$\alpha$ ratio however changes significantly
1350: with abundance, predominantly as a consequence of the larger relative
1351: abundance differences in nitrogen due to its secondary nucleosynthesis
1352: component.
1353: In the shock models, the [\ion{O}{3}]/H$\beta$ does not change
1354: significantly in range either, but the [\ion{N}{2}]/H$\alpha$ ratio
1355: becomes compressed, meaning the diagrams are totally degenerate in
1356: velocity at low metallicity, predominantly due to the changes
1357: photoionized and recombination regions of the shock with metallicity.
1358: The case is similar for the shock+precursor, with the grids becoming
1359: generally more ``compressed'' with decreasing metallicity, and
1360: decreasing the diagnostic power of the diagrams. Note that for each
1361: abundance set the shock-only and shock+precursor models still tend to
1362: be discrete grids, and do not significantly overlap.
1363:
1364: In Figure~\ref{lrat3} we investigate the effect of changing the pre-shock
1365: densities on this \cite{veilleux1987} diagnostic. For the solar abundance
1366: grid, we display the results for the two extreme values of density n$=0.01$
1367: cm$^{-3}$ and n$=1000$ cm$^{-3}$ for both the shock and shock+precursor
1368: models. Note the change in scale between this figure and Figure~\ref{lrat1},
1369: necessary due to the changes in the grids.
1370:
1371: As discussed in section \ref{sec:input}, the models with varying density were
1372: computed not only with the standard magnetic parameter set, but also a range
1373: of magnetic fields to allow for the comparison of both B and B$/\sqrt{\rm n}$
1374: between the model grids. The range of both B and B$/\sqrt{\rm n}$ for these
1375: models are given in Table \ref{param_table}. We include all these models in
1376: Figure~\ref{lrat3} for completeness, meaning that the grids sample more finely
1377: and extend further both the low and high values of the magnetic parameter than
1378: seen in Figures \ref{lrat1}--\ref{lrat2}.
1379:
1380: The low-density model grids are in most respects exactly the same as the n$=1$
1381: cm$^{-3}$ grids seen in Figure \ref{lrat1}, with the only difference arising
1382: due to the extended B$/\sqrt{\rm n}$ range sampled. This similarity is not
1383: surprising, as while the luminosity decreases linearly with the decreased
1384: density, as seen in Equation \ref{Luv_eq}, the ionization parameter is
1385: inversely proportional to density and therefore remains approximately the same
1386: in both the shock recombination region and the precursor region. One
1387: interesting thing to note is how degenerate the grids become upon the
1388: introduction of the very high magnetic parameters.
1389:
1390: However, at high pre-shock densities there are clear changes in both the shock
1391: and shock+precursor grids. In the shock-only grid the low magnetic parameter
1392: models extend to lower values in the both [\ion{O}{3}]/H$\beta$ and
1393: [\ion{N}{2}]/H$\alpha$ ratios by up to 1 dex, while in the shock+precursor
1394: grid only the [\ion{N}{2}]/H$\alpha$ ratio is reduced. This is caused by
1395: collisional de-excitation of the forbidden lines in the photoionized and
1396: recombination regions of the shock, where the electron densities are very
1397: high. The shock+precursor does not change significantly in
1398: [\ion{O}{3}]/H$\beta$ as most of the emission of the [\ion{O}{3}] line arises
1399: from the precursor, which is not high enough density to collisionally
1400: de-excite this transition. There is less of a difference between the grids at
1401: high magnetic parameters and at low to intermediate velocities, because lower
1402: velocities lead to smaller compression factors, and because magnetic pressure
1403: support helps to lessen the degree of gas compression in the tail end of the
1404: shock, meaning that the high density needed to collisionally de-excite these
1405: transitions is not reached.
1406:
1407: \begin{figure}[h]
1408: \includegraphics[scale=0.6]{f32.ps}
1409: \caption{Comparison to SDSS line ratios to the 2$\times$solar abundance model
1410: grids. The thin dotted line represents the \cite{kewley2001} starburst/AGN
1411: classification line, while the thick straight dotted line represents the
1412: Seyfert-LINER division described by \cite{kewley2006}.
1413: \label{sdss_compare2}}
1414: \end{figure}
1415:
1416: Another commonly used \cite{veilleux1987} diagnostic is the
1417: [\ion{O}{3}]$\lambda$5007/H$\beta$ versus
1418: [\ion{S}{2}]$\lambda\lambda$6716,6731/H$\alpha$ diagram. In
1419: Figure~\ref{lrat5} we plot the new grids for this diagram, with the shock and
1420: shock+precursor models plotted separately for clarity, as the overlap between
1421: the grids makes distinguishing them difficult. As with the previous
1422: diagnostic diagram, both the shock and shock+precursor models turn over at
1423: high velocities, becoming twisted and with little separation between the
1424: velocities. Similarly the magnetic parameters show a similar range in spread
1425: in both models, and become more spread with increasing velocity in the
1426: shock-only model. The shape of the shock+precursor grid on this diagram is
1427: again rather different compared to DS95. Like in Figure \ref{lrat4}, the new
1428: shock+precursor model covers a wider range in [\ion{O}{3}]/H$\beta$, with a
1429: lower value at $v_{\rm s}=200$\kms, while it appears somewhat offset to higher
1430: values of [\ion{S}{2}]/H$\alpha$, though not significantly. It also does not
1431: appear to show the degenerate nature see in DS95 (their Figure 2a). The shock models
1432: are harder to compare due to the wider parameter range in the model, but in
1433: general show the same features of increased [\ion{S}{2}]/H$\alpha$ with
1434: velocity, with the turn over around $\sim$500 \kms. The greater range in
1435: magnetic parameter of the new models also means that the spread in
1436: [\ion{O}{3}]/H$\beta$ is much greater at higher velocities as well.
1437:
1438: Figure \ref{lrat6} is a plot of the well-known temperature-sensitive ratio
1439: [\ion{O}{3}]$\lambda$4363/[\ion{O}{3}]$\lambda$5007 (also known as $R_{\rm
1440: OIII}$) versus the [\ion{O}{3}]$\lambda$5007/H$\beta$ ratio, with again the
1441: shock-only and shock+precursor spectra plotted separately. Considering first
1442: the shock-only grid, at low magnetic parameter the electron temperature in the
1443: O$^{2+}$ zone is consistently high, with only the
1444: [\ion{O}{3}]$\lambda$5007/H$\beta$ ratio decreasing with increasing shock
1445: velocity. However, as the magnetic parameter rises, the decreased compression
1446: in the photoionized tail of the shock leads to a higher ionization parameter
1447: in this zone, and greater dominance of [\ion{O}{3}] in the photoionized zone,
1448: which has an electron temperature close to 10000K. As a consequence, the
1449: electron temperature decreases and the [\ion{O}{3}]$\lambda$5007/H$\beta$
1450: ratio increases with increasing shock velocity. Together this leads to the fan
1451: shape observed in the diagram. The shock+precursor grid is dominated by the
1452: bright [\ion{O}{3}] emission of the precursor region, and as the precursor is not
1453: sensitive to the magnetic parameter, this grid covers a reduced area of the parameter
1454: space in the diagram. In terms of velocity, the curve increases
1455: in [\ion{O}{3}]$\lambda$5007/H$\beta$ and decreases in $R_{\rm OIII}$ up to a
1456: velocity of $\sim500$\kms, then turns over and begins to increase in
1457: temperature again. The shock+precursor grid at velocities less than $\sim$450
1458: \kms\ and the high magnetic parameter shock-only grid overlap, meaning this
1459: diagnostic is degenerate in this range. This diagram is very similar to the
1460: DS95, with only the extension to higher velocities and magnetic parameter
1461: range differentiating it. However, this similarity also indicates that the
1462: well-known ``Temperature Problem'', discussed in DS95 and many other papers,
1463: still exists. A comparison of our models with the observations used in DS95 or
1464: larger datasets like SDSS show that the $R_{\rm OIII}$ ratio is too low, and
1465: even different abundance sets are unable to fully solve this issue.
1466:
1467:
1468: \subsection{UV Diagnostics}
1469:
1470: In \cite{allen1998} we emphasized the use of UV line ratio diagrams for the
1471: discrimination between shocks and photoionization models for the NLR. These
1472: diagrams use relatively bright emission lines and reddening-insensitive
1473: ratios. One of the most useful diagrams involves the various ionization stages
1474: of carbon \ion{C}{2}]$\lambda2326$/ \ion{C}{3}] $\lambda$1909
1475: vs.~\ion{C}{4} $\lambda$1550 / \ion{C}{3}] $\lambda$1909. This diagram
1476: is shown in
1477: Figure~\ref{lrat7} with the new solar abundance n$=1$ cm$^{-3}$ shock models.
1478: There is significantly more overlap between the new shock and shock+precursor
1479: grids than with the DS96 models because the higher magnetic parameter shock
1480: models form a fan-like grid on these axes, which almost completely encompasses
1481: the region covered by the shock+precursor models. The shock+precursor grid is
1482: twisted on these axes making it multi-valued at most positions, hence not good
1483: for assessing shock parameters. This diagram does however remain of great use
1484: for separating shock and photoionization models.
1485:
1486: Another useful diagram combines the UV carbon ration of \ion{C}{3}]
1487: $\lambda$1909 / \ion{C}{2}]~$\lambda2326$\ with [\ion{Ne}{3}]~$\lambda$3869 /
1488: [\ion{Ne}{5}]~$\lambda$3426. This has been used by \cite{best2000} and
1489: \cite{Inskip02} to identify shocks and photoionization in the emission line
1490: gas of 3CR and 6C radio galaxies. They show that the ionization state of the
1491: gas varies with radio size such that large radio sources ($>120$ kpc) are
1492: consistent with AGN photoionization, while smaller sources are consistent with
1493: shocks associated with the expanding radio source. Furthermore the extreme gas
1494: kinematics in the smaller radio sources, and detailed consideration of the
1495: energetics and observability of shocks in \cite{Inskip02} supports the
1496: interpretation of shock excitation.
1497:
1498: \begin{figure}[h]
1499: \includegraphics[scale=0.6]{f33.ps}
1500: \caption{Comparison to SDSS line ratios to the 2$\times$ solar abundance model
1501: grids. The thin dotted line represents the \cite{kewley2001} starburst/AGN
1502: classification line, while the thick straight dotted line represents the
1503: Seyfert-LINER division described by \cite{kewley2006}.
1504: \label{sdss_compare3}}
1505: \end{figure}
1506:
1507: As described in \cite{Inskip02}, there is a continuous sequence of objects
1508: which fall in between the shock dominated and photoionization dominated
1509: groups. In these, both mechanisms are likely to play a role and the balance
1510: between shocks and photoionization is may be linked to the radio source
1511: size. As the definition of the shock and photoionization groups relies on the
1512: coverage of the respective model grids, it is important that these grids are
1513: complete. In Figure~\ref{UV_C_Ne_diag} we reconstruct the same \ion{C}{3}]~
1514: $\lambda$1909 / \ion{C}{2}]~$\lambda2326$ versus
1515: [\ion{Ne}{3}]~$\lambda$3869/ [\ion{Ne}{5}]~$\lambda$3426 diagram as used by
1516: \cite{best2000} and \cite{Inskip02} using the data tabulated in those
1517: papers. We overlay the new shock model grids to demonstrate how these
1518: relate to, and possible help explain, these previously-made
1519: conclusions. For comparison, we also plot the set of new and updated
1520: AGN photoionization models from \cite{groves2004b} as well as the
1521: A$_{M/I}$ models from \cite{bws96} that combine matter- and
1522: ionization-bounded clouds to create the observed sequence.
1523:
1524: The shock and shock+precursor models are those for new solar abundance n$=1$
1525: cm$^{-3}$. As in the previous diagram, the shock models define a fan shaped
1526: grid, and the shock+precursor grid turns over in both ratios. With the
1527: wider range in shock velocity and magnetic parameter the new models cover a
1528: greater region of this diagram than the DS96 models used by \cite{best2000}
1529: and \cite{Inskip02}. Indeed a number of 6C sources which have \ion{C}{3}] /
1530: $\lambda$1909 \ion{C}{2}]$\lambda2326$\ ratios intermediate between the main
1531: shock and photoionization groups, fall within the higher velocity
1532: shock+precursor grid (6C~1017, 6C~1256, 6C~0943, 6C~1019).
1533:
1534: The dust-free and dusty photoionization models
1535: from \cite{groves2004b} demonstrate the other extreme of ionization,
1536: where the emission arises totally from gas excited by the ionizing
1537: radiation emitted by the accretion of gas onto the central black hole
1538: of the AGN. The dust-free models represent \mapiii\
1539: photoionization models that have been calculated for a sequence of ionization
1540: parameters and densities. The 1Z$_{\odot}$, $\alpha$=-1.4, n$_{\rm
1541: H}$=10$^2$-10$^4$ models shown here behave in a similar way to the
1542: \mapii\ models described in \cite{allen1998} (and used by
1543: \cite{best2000} and \cite{Inskip02}). The dusty models incorporate the effect
1544: of radiation pressure and result in a stagnation of the ionization parameter
1545: at high values, offering an explanation for the similarity of Seyfert NLR
1546: spectra. Figure~\ref{UV_C_Ne_diag} shows that the dusty photoionization models
1547: can produce line ratios as observed in the larger 3CR and 6C radio
1548: sources. The dusty photoionization model plotted here is the 1Z$_{\odot}$,
1549: $\alpha$=-1.4, n$_{\rm H}$=10$^2$ and n$_{\rm H}$=10$^4$ model, and a
1550: discussion of these models in terms of the \cite{best2000} and
1551: \cite{Inskip02} data, as well as a full
1552: grid of such models on the same diagram can be found in
1553: \cite{groves2004b}.
1554: %, specifically section 3.2 and Figure~5.
1555:
1556: \subsection{IR Diagnostics}
1557:
1558: The mid- and far-IR emission from galaxies is dominated by the emission of
1559: dust that is heated by UV radiation. IR spectra contain PAH and dust features
1560: and are also rich in atomic fine structure emission lines. These spectral
1561: features, as observed by ISO and Spitzer offer a wealth of information for
1562: studying the nature of the circumnuclear dust, and the contributions of the
1563: AGN, shocks and starbursts to the total IR emission.
1564:
1565: \cite{genzel1998} introduced infrared diagnostic diagrams using ratios of
1566: [\ion{Ne}{5}]14.3$\mu$m / [\ion{Ne}{2}]12.8$\mu$m and
1567: [\ion{O}{4}]25.9$\mu$m / [\ion{Ne}{2}]12.8$\mu$m to investigate starbursts,
1568: ULIRGs and AGNs . They used these line ratios, combined with PAH strengths to
1569: place limits on the percentage contribution of AGN and starburst contributions
1570: to the IR emission of ULIRGs. Mid-IR diagnostics for LINERs are described by
1571: \cite{sturm2006}, where they use the high ionization [\ion{O}{4}]25.9$\mu$m
1572: and [\ion{Ne}{5}]14.3$\mu$m lines to identify the presence of AGN in
1573: $\sim90$\% of LINERs. They confirm the differences in properties between
1574: IR-faint and IR-luminous LINERs, and identify the need to disentangle the
1575: various stellar, \ion{H}{2} region and AGN processes at work in these objects.
1576:
1577:
1578:
1579:
1580: \cite{groves2006} investigated the IR emission of the NLR distinct from the
1581: emission of the torus, showing that the NLR emission can contribute up to
1582: $\sim10$\% of the IRAS 25$\mu$m flux. They emphasize the fact that high
1583: ionization lines like [\ion{Ne}{5}]14.3$\mu$m arises only in the NLR, and
1584: their diagnostic diagram which utilizes only IR lines of neon provides a very
1585: useful indicator of AGN and starburst contributions.
1586:
1587: Here we present a set of IR diagnostic diagrams drawn from these previous
1588: works, overlaid with the solar abundance shock and shock + precursor model
1589: grids. Figure~\ref{Ne_IR_diag}
1590: ([\ion{Ne}{5}]14.3$\mu$m / [\ion{Ne}{2}]12.8$\mu$m versus
1591: [\ion{Ne}{3}]15.5$\mu$m / [\ion{Ne}{2}]12.8$\mu$m) combines three different
1592: ionization stages of neon removing any abundance dependence. Figure
1593: \ref{Ne_O_IR_diag}, [\ion{Ne}{5}]14.3$\mu$m / [\ion{Ne}{2}]12.8$\mu$m versus
1594: [\ion{O}{4}]25.9$\mu$m / [\ion{Ne}{2}]12.8$\mu$m, utilizes two high ionization
1595: species of \ion{O}{4} and \ion{Ne}{5} whose ratios to \ion{Ne}{2} provide
1596: strong discriminants of AGN versus starburst processes. Figure ~\ref{S_Ne}
1597: displays the model results for [\ion{S}{4}]10.44$\mu$m / [\ion{S}{3}]18.7$\mu$m
1598: versus [\ion{Ne}{3}]15.5$\mu$m / [\ion{Ne}{2}]12.8$\mu$m.
1599:
1600: \cite{lutz2003} used the ratios of
1601: [\ion{Fe}{2}]26.0$\mu$m / [\ion{O}{4}]25.9$\mu$m versus
1602: [\ion{O}{4}]25.9$\mu$m / [\ion{Ne}{2}]12.8$\mu$m to distinguish between gas
1603: ionized by early-type stars, AGN and shocks. \cite{sturm2006} showed how the
1604: same diagram separates starburst galaxies, Seyfert galaxies and supernova
1605: remnants. Figure~\ref{lutz_fig} shows how the shock and shock + precursor
1606: model grids form relatively tight shock velocity sequences in this
1607: diagram. The slope of these sequences is similar to the distribution of
1608: observed ratios in \cite{sturm2006} but the models predict systematically
1609: higher ratios than observed, except for some LINER objects which fall in the
1610: region of the lower velocity shock model grid.
1611:
1612: \subsection{Comparison with SDSS Observations}
1613: We now compare the new shock and shock+precursor models to emission line
1614: ratios of AGN and star-forming galaxies observed in the Sloan Digital Sky
1615: Survey (SDSS). To do this we have adopted the sample of narrow emission line
1616: galaxies compiled by \cite{hao2005}. The classification of the sample into
1617: broad- and narrow-line AGN, and star-forming galaxies is considered in detail
1618: in \cite{hao2005}. Here we choose to use the complete narrow emission line
1619: galaxies sample of $\sim 42000$ sources, and use the AGN/star-forming galaxy
1620: separation schemes of \cite{kewley2001}, \cite{kauffmann2003}, and
1621: \cite{kewley2006} in order to
1622: emphasize where our model grids lie with respect to these different classes of
1623: objects.
1624:
1625: \begin{figure}[h]
1626: \includegraphics[scale=0.6]{f34.ps} \\
1627: \caption{Comparison to SDSS line ratios to an oxygen excitation diagnostic in
1628: which the regions identified as being excited by \ion{H}{2} regions, Seyferts
1629: and LINERs are labelled. Note that shock-only spectra may characterize some
1630: LINERs, but that the Seyfert galaxies are not well fit by shock+precursor
1631: models on this particular diagram.
1632: \label{sdss_compare4}}
1633: \end{figure}
1634:
1635: Figure \ref{sdss_compare1} shows the first of the familiar
1636: \citet{veilleux1987} diagnostics; [\ion{O}{3}]$\lambda$5007/H$\beta$
1637: versus [\ion{N}{2}]$\lambda$6583/H$\alpha$. We have overlaid as a
1638: density plot the line ratios observed in the SDSS the narrow emission line
1639: sample.
1640:
1641: The distribution of emission line galaxies shows two main branches. The star
1642: forming galaxy branch sweeps in a curve showing relatively small scatter from
1643: upper left to lower right. This is largely an abundance sequence, with
1644: abundances increasing towards the lower right \citep{dopita2000,kewley2001b,
1645: Dopita06}. The AGN are mostly distributed in an arm which extends from the
1646: base of the star-forming sequence up towards the upper right of the
1647: diagram. AGN generally have higher values of
1648: [\ion{N}{2}]$\lambda$6583/H$\alpha$, and the distribution extends to higher
1649: values of [\ion{O}{3}]$\lambda$5007/H$\beta$. The dotted and dashed curved
1650: lines represent the \cite{kewley2001} and \cite{kauffmann2003} classification
1651: lines respectively. The \cite{kewley2001} classification is based on the
1652: theoretical maximum line ratios possible by pure stellar photoionization.
1653: Galaxies above this line are most likely dominated by AGN. The
1654: \cite{kauffmann2003} line is a purely empirical dividing line between pure
1655: star-forming galaxies, and Seyfert-H{\rm II} composite objects.
1656:
1657: For comparison, we show in Figure \ref{sdss_compare1} the shock and
1658: shock+precursor grids for the $2\times$solar metallicity models. This choice
1659: of metallicity is driven by the work of \cite{groves2004} and
1660: \cite{groves2006}, who find that super-solar metallicity photoionization
1661: models best reproduce the observed narrow-line ratios in AGN, and the work of
1662: \cite{kauffmann2003}, who find that AGN are typically hosted in galaxies more
1663: massive than $10^{10}$ M$_{\odot}$, and therefore likely to contain high
1664: metallicity gas \citep[e.g.][]{tremonti04}
1665:
1666: Figure \ref{sdss_compare1} reveals that both the shock and shock+precursor
1667: models are generally located in the AGN region of this diagram, lying mostly
1668: above the \cite{kewley2001} classification curve. The shock+precursor grid
1669: overlaps well with the strong AGN or Seyfert branch, lying above the
1670: \cite{kewley2006} LINER/Seyfert dividing line, and extends along the branch
1671: with increasing shock velocity. The high shock velocity portion of the grid
1672: extends roughly to the limit of the observed distribution before folding over
1673: on itself at the highest velocities. At lower velocities, the shock+precursor
1674: models extend into the ``composites'' region and even into the starformation
1675: or H{\sc II} region below the \cite{kauffmann2003} curve at the lowest
1676: velocities and magnetic parameters.
1677:
1678: The lower velocity shock-only models also overlap the AGN branch, but are
1679: predominantly located in the LINER region, simultaneously extending out and
1680: spreading out (with magnetic parameter) at the higher shock velocities.
1681:
1682: The emission-line galaxies form similar distributions on the other two
1683: \citet{veilleux1987} diagnostic diagrams; [\ion{O}{3}]$\lambda$5007/H$\beta$
1684: versus [\ion{S}{2}]$\lambda\lambda$6716,6731/H$\alpha$, and the plot of
1685: [\ion{O}{3}]$\lambda$5007/H$\beta$ versus [\ion{O}{1}]$\lambda$6300/H$\alpha$
1686: (see Figures \ref{sdss_compare2} and \ref{sdss_compare3}). The
1687: \cite{kewley2001} division between AGN and \ion{H}{2} region excitation is
1688: shown as dotted lines on these figures.
1689:
1690: In these figures, the AGN branches show a bifurcation, that has been used by
1691: \cite{kewley2006} to distinguish between Seyfert and LINER galaxies, and so
1692: provide a general classification scheme for AGN host galaxies. The
1693: Seyfert-LINER dividing line is shown in Figures \ref{sdss_compare2} and
1694: \ref{sdss_compare3} as a thick dotted line. The shock and shock+precursor
1695: grids overlap the AGN distributions, but tend to fall in the region of the
1696: Seyfert-LINER division. The shock-only models mostly fall in the LINER region
1697: of these diagnostic plots, but tend to extend to extend to higher values of
1698: [\ion{S}{2}]$\lambda\lambda$6716,6731/H$\alpha$, and
1699: [\ion{O}{1}]$\lambda$6300/H$\alpha$ than is observed. There is also more
1700: overlap with the shock+precursor models than in seen in
1701: Figure~\ref{sdss_compare1}.
1702:
1703: Figure~\ref{sdss_compare4} shows the
1704: [\ion{O}{3}]$\lambda$5007/[\ion{O}{2}]$\lambda\lambda$3726,3729 versus
1705: [\ion{O}{1}]$\lambda$6300/H$\alpha$ diagnostic diagram over-plotted with the
1706: classification scheme from \cite{kewley2006}. This diagram provides the
1707: cleanest separation between the two AGN branches. In photoionized plasmas, the
1708: [\ion{O}{3}]$\lambda$5007/[\ion{O}{2}]$\lambda\lambda$3726,3729 is sensitive
1709: to the specific intensity of the radiation field, and the
1710: [\ion{O}{1}]$\lambda$6300/H$\alpha$ ratio to the hardness or spectral index of
1711: the radiation field. As pointed out by \cite{kewley2006} this diagram provides
1712: a simple method for classification, but is more sensitive to reddening
1713: correction of the
1714: [\ion{O}{3}]$\lambda$5007/[\ion{O}{2}]$\lambda\lambda$3726,3729 ratio. Once
1715: again, the shock+precursor grids fall predominantly close to the Seyfert-LINER
1716: dividing line, and so describe neither the Seyferts or the LINERs particularly
1717: well. The pure shock models are a much better fit to the LINER sequence,
1718: provided that the shock velocities are not too great. The highest shock
1719: velocities have too strong [\ion{O}{1}]$\lambda$6300/H$\alpha$ ratios.
1720:
1721: In conclusion, Seyfert galaxies are not well described by the shock +
1722: precursor models. For these, the radiation-pressure dominated photoionization
1723: models \citep{dopita2002, groves2004a, groves2004b} provide a much better
1724: description of the spectra. The LINERs, on the other hand, fit better to shock
1725: only models, and it is likely that at least some of these objects are in fact
1726: shock-excited. A good example of a LINER which is known to be shock-excited is
1727: the nuclear disk of M87 \citep{dopita1997}.
1728:
1729: \section{The On-line Library}
1730: \label{online}
1731:
1732: The complete electronic files that comprise the \mapiii\ Library of Fast Shock
1733: Models are available via the Shock Model Portal of the MAPPINGS online web
1734: pages at \url{http://cdsweb.u-strasbg.fr/\~{}allen/shock.html}.
1735: The original \mapiii\ output files are available for each model in the
1736: library. We also provide tables of emission line ratios, and column
1737: densities. These files are organized into the various model sets characterized
1738: by a given chemical abundance and pre-shock density and then into velocity
1739: sequences for a given magnetic field. The files are listed on the online pages
1740: using the model names shown in Table \ref{param_table}.
1741:
1742: The emission line ratio tables are wavelength ordered lists of the flux ratios
1743: of all the emission lines calculated in \mapiii, and are given with respect to
1744: H$\beta$=1. Each table contains the ratios for a velocity sequence of models,
1745: 100-1000 km\ s$^{-1}$, for a given abundance set, density and magnetic
1746: field. Shock and precursor components are tabulated separately and we also
1747: provide tables of the emission line ratios for the combination of
1748: shock+precursor. Each emission line ratio table also includes the absolute
1749: luminosity of the H$\beta$ line in units of Log$_{10}$(erg
1750: cm$^{-2}$~s$^{-1}$).
1751:
1752: The column density tables contain the integrated model column densities for
1753: all of the ionic stages of each of the elements in the corresponding abundance
1754: list. Each table includes the column densities for a velocity sequence of
1755: models in units of cm$^{-2}$. Column densities for the shock and precursor
1756: components are tabulated separately.
1757:
1758: In addition to the tabulated model data, we also provide programs for
1759: accessing the library of models, and for generating various plots. These
1760: programs are coded using IDL\footnote{\url{http://www.ittvis.com/idl/}} and
1761: can be used as interactive graphical user interface widgets, and also via the
1762: IDL command line. The SHOCKPLOT package allows plotting the models on 2-D line
1763: ratio diagrams using any linear combination of line ratios. SHOCKPLOT enables
1764: quick browsing through the many grids of models, and can also be used as a
1765: procedure call from other IDL programs to over-plot model grids on observed
1766: line ratio data.
1767:
1768: \section{Summary}
1769:
1770: We have presented an extensive library of radiative shock models covering a
1771: wide range of shock velocities, magnetic fields, densities and abundances. The
1772: shock model predictions for the ionizing radiation, temperatures, and
1773: luminosities generated by shocks will be applicable in a wide range of
1774: astrophysical situations. The solar abundance models supersede the models of
1775: DS96, with model code improvements leading to some differences in the
1776: predictions of nitrogen species, and also for models with very low magnetic
1777: parameter. Differences in the input abundances result in the most significant
1778: variations between the grids of models, and the range and sampling of the
1779: shock velocity and magnetic field provides much a more detailed and complete
1780: set of model predictions than previously available. The extension to higher
1781: shock velocities is important for the analysis of emission line regions of
1782: active galaxies where shocks of this speed are expected in jet-cloud
1783: interactions \citep{saxton2005}. At these shock velocities some of the
1784: commonly used line ratios such as [\ion{O}{3}]$\lambda$5007/H$\beta$ turn over
1785: and do not follow simple extrapolations from the DS96 model grids.
1786:
1787: Included as a part of this library, the physical structure of the
1788: shocks and their precursors give insight into how the physical parameters
1789: of the shock lead to the resulting continuum and line emission. We
1790: have included in this paper examples of these, exploring the range of
1791: the parameters and demonstrating the effects of these of the density,
1792: temperature and ionization structure.
1793:
1794: We have presented the model grids on a set of standard UV, optical and IR line
1795: ratio diagrams, and we have compared of the new models to the example data
1796: sets of radio galaxies and the SDSS sample of narrow emission line
1797: galaxies. The updated version of the \ion{C}{3}]$\lambda$1909 /
1798: [\ion{C}{2}]$\lambda2326$ vs.~[\ion{Ne}{3}]$\lambda$3869 /
1799: [\ion{Ne}{5}]$\lambda$3426 diagram used by \cite{best2000} and \cite{Inskip02}
1800: for the analysis of emission line regions of radio galaxies, shows that some
1801: of the 6C sources fall within the new higher velocity shock+precursor
1802: grid. This supports the interpretation that the emission line regions in
1803: smaller radio sources are excited by shocks. The new library of models
1804: presented here, and a more complete set of emission line observations of these
1805: sources will allow a much more detailed analysis of the contributions of shock
1806: and photoionization in these sources.
1807:
1808: Large samples of emission line ratios such as now available from the SDSS
1809: provide an extremely valuable resource for classifying and analyzing the
1810: emission line excitation mechanisms in active and star-forming galaxies. The
1811: comparison of the narrow emission line galaxies sample (compiled by
1812: \cite{hao2005}) to the shock models shows that shocks do predict line ratios
1813: in the observed range, and that shocks may indeed provide the best explanation
1814: for LINER spectra. This library of shock models, combined with detailed grids
1815: of (AGN and star-formation) photoionization models now available should allow a
1816: new statistical approach to analyzing the contributions of shocks,
1817: star-formation and AGN photoionization to excitation of emission lines in
1818: galaxies.
1819:
1820: The complete set of electronic files that comprise the library are available
1821: on-line, along with tools to assist in the comparison of observations to the
1822: model predictions line ratio diagrams. Together this library provides
1823: one of the largest databases of radiative shock models and a unique
1824: tool in the interpretation and diagnosis of fast shocks.
1825:
1826:
1827: \acknowledgments We thank the anonymous referee for comments which have improved the
1828: clarity of this paper. BG would like to thank the Observatoire Astronomique de
1829: Strasbourg for their financial support and hospitality. MAD acknowledges the
1830: support of both the Australian National University and of the Australian
1831: Research Council (ARC) through his (2002-2006) ARC Australian Federation
1832: Fellowship, and also under the ARC Discovery projects DP0208445 and
1833: DP0342844. Both Dopita and Sutherland acknowledge support under the ARC
1834: Discovery project DP0664434.
1835:
1836:
1837: \begin{thebibliography}{}
1838:
1839: \bibitem[Allen et al.(1998)]{allen1998} Allen, M.~G., Dopita, M.~A., \&
1840: Tsvetanov, Z.~I.\ 1998, \apj, 493, 571
1841:
1842: \bibitem[Allen et al.(1999)]{allen1999} Allen, M.~G., Dopita, M.~A.,
1843: Tsvetanov, Z.~I., \& Sutherland, R.~S.\ 1999, \apj, 511, 686
1844:
1845: \bibitem[Anders \& Grevesse(1989)]{Anders89}Anders, E. \& Grevesse, N. 1989,
1846: Geochimica et Cosmochimica Acta, 53, 197
1847:
1848: \bibitem[Asplund, Grevesse \& Sauval(2005)]{Asplund05}Asplund, M., Grevesse,
1849: N., \& Sauval, A.~J. 2005, in ASP Conference Series, Vol. 336 "Cosmic
1850: Abundances as Records of Stellar Evolution and Nucleosynthesis", eds. Thomas
1851: G. Barnes III and Frank N. Bash. San Francisco: Astronomical Society of the
1852: Pacific, p.25
1853:
1854: \bibitem[Baldwin et al.(1981)]{baldwin1981} Baldwin, J.~A., Phillips, M.~M.,
1855: \& Terlevich, R.\ 1981, \pasp, 93, 5
1856:
1857: \bibitem[Best et al.(2000)]{best2000} Best, P.~N., R{\" o}ttgering, H.~J.~A.,
1858: \& Longair, M.~S.\ 2000, \mnras, 311, 23
1859:
1860: \bibitem[Binette et al.(1996)]{bws96} Binette, L., Wilson, A.~S., \&
1861: Storchi-Bergmann, T.\ 1996, \aap, 312, 365
1862:
1863:
1864: \bibitem[Bradley et al.(2004)]{bradley2004} Bradley, L.~D., Kaiser, M.~E., \&
1865: Baan, W.~A.\ 2004, \apj, 603, 463
1866:
1867: \bibitem[De Breuck et al.(2000)]{debreuck2000} De Breuck, C., R{\" o}ttgering,
1868: H., Miley, G., van Breugel, W., \& Best, P.\ 2000, \aap, 362, 519
1869:
1870: \bibitem[Dopita(1977)]{Dopita77} Dopita, M.~A. 1977, \apjs, 56, 437
1871:
1872:
1873: \bibitem[Dopita \& Sutherland(1995)]{dopita1995} Dopita, M.~A., \& Sutherland,
1874: R.~S.\ 1995, \apj, 455, 468 (DS95)
1875:
1876: \bibitem[Dopita \& Sutherland(1996)]{dopita1996} Dopita, M.~A., \& Sutherland,
1877: R.~S.\ 1996, \apjs, 102, 161 (DS96)
1878:
1879: \bibitem[Dopita et al.(1997)]{dopita1997} Dopita, M.~A., Koratkar, A.~P.,
1880: Allen, M.~G., Tsvetanov, Z.~I., Ford, H.~C., Bicknell, G.~V., \& Sutherland,
1881: R.~S.\ 1997, \apj, 490, 202
1882:
1883: \bibitem[Dopita et al.(2000)]{dopita2000} Dopita, M.~A.,Kewley, L.~J., \&
1884: Sutherland, R.~S. 2000, \apj, 542, 224
1885:
1886: \bibitem[Dopita et al.(2002)]{dopita2002} Dopita, M.~A., \& Groves, B.~A.,
1887: Sutherland, R.~S., Binette, L. \& Cecil, G. 2002, \apj, 572, 753
1888:
1889: \bibitem[Dopita \& Sutherland(2003)]{dopita2003} Dopita, M.~A., \& Sutherland,
1890: R.~S.\ 2003, Astrophysics of the diffuse universe, Berlin, New York: Springer,
1891: 2003.~Astronomy and astrophysics library, ISBN 3540433627
1892:
1893: \bibitem[Dopita et al.(2005)]{dopita2005} Dopita, M.~A., et al.\ 2005, \apj,
1894: 619, 755
1895:
1896: \bibitem[Dopita et al.(2006)]{Dopita06} Dopita, M.~A., et al.\ 2006, \apjs,
1897: 167, 177
1898:
1899: \bibitem[Draine \& Salpeter(1979)]{Draine79}Draine, B.~T., \& Salpeter,
1900: E.~E. 1979, \apj, 231, 77
1901:
1902: \bibitem[Dwek, Foster \& Vancura(1996)]{Dwek92}Dwek, E., Foster, S.~M., \&
1903: Vancura, O. 1996, \apj, 457, 244
1904:
1905: \bibitem[Evans et al.(1999)]{evans1999} Evans, I., Koratkar, A., Allen, M.,
1906: Dopita, M., \& Tsvetanov, Z.\ 1999, \apj, 521, 531
1907:
1908: \bibitem[Feinstein et al.(1999)]{feinstein1999} Feinstein, C., Macchetto,
1909: F.~D., Martel, A.~R., Sparks, W.~B., \& McCarthy, P.~J.\ 1999, \apj, 526, 623
1910:
1911: \bibitem[Ferruit et al.(1999)]{ferruit1999} Ferruit, P., Wilson, A.~S.,
1912: Whittle, M., Simpson, C., Mulchaey, J.~S., \& Ferland, G.~J.\ 1999, \apj, 523,
1913: 147
1914:
1915: \bibitem[Fox et al.(2004)]{fox2004} Fox, A.~J., Savage, B.~D., Wakker, B.~P.,
1916: Richter, P., Sembach, K.~R., \& Tripp, T.~M.\ 2004, \apj, 602, 738
1917:
1918: \bibitem[Genzel et al.(1998)]{genzel1998} Genzel, R., et al.\ 1998, \apj, 498,
1919: 579
1920:
1921: \bibitem[Groves et al.(2006)]{groves2006} Groves, B., Dopita, M., \&
1922: Sutherland, R.\ 2006, \aap, 458, 405
1923:
1924:
1925: \bibitem[Groves et al.(2004)]{groves2004} Groves, B.~A., Cecil, G., Ferruit,
1926: P., \& Dopita, M.~A.\ 2004, \apj, 611, 786
1927:
1928: \bibitem[Groves et al.(2004a)]{groves2004a} Groves, B.~A., Dopita, M.~A., \&
1929: Sutherland, R.~S.\ 2004a, \apjs, 153, 9
1930:
1931: \bibitem[Groves et al.(2004b)]{groves2004b} Groves, B.~A., Dopita, M.~A., \&
1932: Sutherland, R.~S.\ 2004b, \apjs, 153, 75
1933:
1934: \bibitem[Groves et al.(2007)]{groves2007} Groves, B.~A., et al. in
1935: preparation.
1936:
1937: \bibitem[Hao et al.(2005)]{hao2005} Hao, L., et al.\ 2005, \aj, 129, 1783
1938:
1939: \bibitem[Heng et al.(2007)]{heng2007} Heng, K., van Adelsberg,
1940: M., McCray, R., \& Raymond, J.~C.\ 2007, \apj, 668, 275
1941:
1942: \bibitem[Inskip et al.(2002)]{Inskip02} Inskip, K.~J., Best, P.~N., Rawlings,
1943: S., Longair, M.~S., Cotter, G., R{\" o}ttgering, H.~J.~A., \& Eales, S.\ 2002,
1944: \mnras, 337, 1381
1945:
1946: \bibitem[Iwamuro et al.(2003)]{iwamuro2003} Iwamuro, F., et al.\ 2003, \apj,
1947: 598, 178
1948:
1949: \bibitem[Jones et al.(1994)]{Jones94}Jones, A.~P., Tielens, A.~G.~G.~M.,
1950: Hollenbach, D.~J., \& McKee, C.~F. 1994, \apj, 433, 797
1951:
1952: \bibitem[Jones et al.(1996)]{Jones96}Jones, A.~P., Tielens, A.~G.~G.~M., \&
1953: Hollenbach, D.~J. 1996, \apj, 469, 740
1954:
1955: \bibitem[Kauffmann et al.(2003)]{kauffmann2003} Kauffmann, G., et al.\ 2003,
1956: \mnras, 346, 1055
1957:
1958: \bibitem[Kewley et al.(2001a)]{kewley2001} Kewley, L.~J., Dopita, M.~A.,
1959: Sutherland, R.~S., Heisler, C.~A., \& Trevena, J.\ 2001, \apj, 556, 121
1960:
1961: \bibitem[Kewley et al.(2001b)]{kewley2001b} Kewley, L.~J., Heisler, C.~A.,
1962: Dopita, M.~A., \& Lumsden, S. \ 2001, \apjs, 132, 37
1963:
1964: \bibitem[Lutz et al.(2003)]{lutz2003} Lutz, D., Sturm, E., Genzel, R., Spoon,
1965: H.~W.~W., Moorwood, A.~F.~M., Netzer, H., \& Sternberg, A.\ 2003, \aap, 409,
1966: 867
1967:
1968: \bibitem[Kewley et al.(2006)]{kewley2006} Kewley, L.~J., Groves, B.,
1969: Kauffmann, G., \& Heckman, T.\ 2006, \mnras, 372, 961
1970:
1971: \bibitem[Leitherer et al.(1996)]{leitherer1996} Leitherer, C., et al.\ 1996,
1972: \pasp, 108, 996
1973:
1974: \bibitem[McKee \& Hollenbach(1980)]{mckee1980} McKee, C.~F., \& Hollenbach,
1975: D.~J.\ 1980, \araa, 18, 219
1976:
1977: \bibitem[Maxfield et al.(2002)]{maxfield2002} Maxfield, L., Spinrad, H.,
1978: Stern, D., Dey, A., \& Dickinson, M.\ 2002, \aj, 123, 2321
1979:
1980: \bibitem[Nelson et al.(2000)]{nelson2000} Nelson, C.~H., Weistrop, D.,
1981: Hutchings, J.~B., Crenshaw, D.~M., Gull, T.~R., Kaiser, M.~E., Kraemer, S.~B.,
1982: \& Lindler, D.\ 2000, \apj, 531, 257
1983:
1984: \bibitem[Perna \& Lazzati(2002)]{perna2002} Perna, R., \& Lazzati, D.\ 2002, \apj, 580, 261
1985:
1986: \bibitem[Pineau de F\^orets \& Flower(1997)]{Pineau97}Pineau de F\^orets,
1987: G. \& Flower, D. 1997, in IAU Symp. 178, \emph{Molecules in Astrophysics:
1988: Probes \& Processes}, ed. E.~F. van Dishoeck, (Kluwer:Dordrecht), p113
1989:
1990: \bibitem[Raymond(1979)]{Raymond79}Raymond, J.~C. 1979, \apjs, 39, 1
1991:
1992: \bibitem[Reuland et al.(2007)]{Reuland07}Reuland, M.~A. et al. 2007, \aj, 133,
1993: 2607
1994:
1995: \bibitem[Russell \& Dopita(1992)]{Russell92}Russell, S.~C. \& Dopita,
1996: M.~A. 1992, \apj, 384, 508
1997:
1998: \bibitem[Sabra et al.(2003)]{sabra2003} Sabra, B.~M., Shields, J.~C., Ho,
1999: L.~C., Barth, A.~J., \& Filippenko, A.~V.\ 2003, \apj, 584, 164
2000:
2001: \bibitem[Saxton et al.(2005)]{saxton2005} Saxton, C.~J., Bicknell, G.~V.,
2002: Sutherland, R.~S., \& Midgley, S.\ 2005, \mnras, 359, 781
2003:
2004:
2005: \bibitem[Shull et al.(2003)]{shull2003} Shull, J.~M., Tumlinson, J., \&
2006: Giroux, M.~L.\ 2003, \apjl, 594, L107
2007:
2008: \bibitem[Sturm et al.(2006)]{sturm2006} Sturm, E., et al.\ 2006, \apjl, 653,
2009: L13
2010:
2011: \bibitem[Sutherland et al.(1993)]{sutherland1993} Sutherland, R.~S., Bicknell,
2012: G.~V., \& Dopita, M.~A.\ 1993, \apj, 414, 510
2013:
2014: \bibitem[Sutherland \& Dopita(1993)]{sutherland1993b} Sutherland, R.~S., \&
2015: Dopita, M.~A.\ 1993, \apjs, 88, 253
2016:
2017: \bibitem[Sutherland et al.(2003)]{sutherland2003a} Sutherland, R.~S., Bisset,
2018: D.~K., \& Bicknell, G.~V.\ 2003, \apjs, 147, 187
2019:
2020: \bibitem[Sutherland et al.(2003)]{sutherland2003b} Sutherland, R.~S.,
2021: Bicknell, G.~V., \& Dopita, M.~A.\ 2003, \apj, 591, 238
2022:
2023: \bibitem[Taniguchi et al.(2001)]{taniguchi2001} Taniguchi, Y., et al.\ 2001,
2024: \apjl, 559, L9
2025:
2026: \bibitem[Tremonti et al.(2004)]{tremonti04} Tremonti, C.~A., et
2027: al.\ 2004, \apj, 613, 898
2028:
2029: \bibitem[Veilleux \& Osterbrock(1987)]{veilleux1987} Veilleux, S., \&
2030: Osterbrock, D.~E.\ 1987, \apjs, 63, 295
2031:
2032: \bibitem[Villar-Martin et al.(1997)]{villar1997} Villar-Martin, M., Tadhunter,
2033: C., \& Clark, N.\ 1997, \aap, 323, 21
2034:
2035: \bibitem[Villar-Mart{\'{\i}}n et al.(2003)]{villar2003}
2036: Villar-Mart{\'{\i}}n, M., Vernet, J., di Serego Alighieri, S., Fosbury, R.,
2037: Humphrey, A., \& Pentericci, L.\ 2003, \mnras, 346, 273
2038:
2039: \bibitem[Whittle et al.(2005)]{whittle2005} Whittle, M., Rosario, D.~J.,
2040: Silverman, J.~D., Nelson, C.~H., \& Wilson, A.~S.\ 2005, \aj, 129, 104
2041:
2042: \end{thebibliography}
2043:
2044: \clearpage
2045:
2046: \LongTables
2047: \begin{deluxetable}{lrr}
2048: \tablecolumns{3}
2049: \tablewidth{0pc}
2050: \tabletypesize{\small}
2051: \tablecaption{Model Parameters \label{param_table}}
2052: \tablehead{
2053: \colhead{Model name} &
2054: \colhead{B ($\mu$G)} &
2055: \colhead{B/$\sqrt{\rm n}$ ($\mu$G\ cm$^{3/2}$)} }
2056: \startdata
2057: \cutinhead{Solar abundance, n=1.0, v=100,125,...1000}
2058: M\_n1\_b0 & 1.0e-04 & 1.0e-04 \\
2059: M\_n1\_b0.5 & 0.50 & 0.50 \\
2060: M\_n1\_b1 & 1.00 & 1.00 \\
2061: M\_n1\_b2 & 2.00 & 2.00 \\
2062: M\_n1\_be & 3.23 & 3.23 \\
2063: M\_n1\_b4 & 4.00 & 4.00 \\
2064: M\_n1\_b5 & 5.00 & 5.00 \\
2065: M\_n1\_b10 & 10.0 & 10.0 \\
2066: \cutinhead{Solar $\times$2 abundance, n=1.0, v=100,125,...1000}
2067: R\_n1\_b0 & 1.0e-04 & 1.0e-04 \\
2068: R\_n1\_b0.5 & 0.50 & 0.50 \\
2069: R\_n1\_b1 & 1.00 & 1.00 \\
2070: R\_n1\_b2 & 2.00 & 2.00 \\
2071: R\_n1\_be & 3.23 & 3.23 \\
2072: R\_n1\_b4 & 4.00 & 4.00 \\
2073: R\_n1\_b5 & 5.00 & 5.00 \\
2074: R\_n1\_b10 & 10.0 & 10.0 \\
2075: \cutinhead{Dopita 2005 abundance, n=1.0, v=100,125,...1000}
2076: J\_n1\_b0 & 1.0e-04 & 1.0e-04 \\
2077: J\_n1\_b0.5 & 0.50 & 0.50 \\
2078: J\_n1\_b1 & 1.00 & 1.00 \\
2079: J\_n1\_b2 & 2.00 & 2.00 \\
2080: J\_n1\_be & 3.23 & 3.23 \\
2081: J\_n1\_b4 & 4.00 & 4.00 \\
2082: J\_n1\_b5 & 5.00 & 5.00 \\
2083: J\_n1\_b10 & 10.0 & 10.0 \\
2084: \cutinhead{SMC abundance, n=1.0, v=100,125,...1000}
2085: P\_n1\_b0 & 1.0e-04 & 1.0e-04 \\
2086: P\_n1\_b0.5 & 0.50 & 0.50 \\
2087: P\_n1\_b1 & 1.00 & 1.00 \\
2088: P\_n1\_b2 & 2.00 & 2.00 \\
2089: P\_n1\_be & 3.23 & 3.23 \\
2090: P\_n1\_b4 & 4.00 & 4.00 \\
2091: P\_n1\_b5 & 5.00 & 5.00 \\
2092: P\_n1\_b10 & 10.0 & 10.0 \\
2093: \cutinhead{LMC abundance, n=1.0, v=100,125,...1000}
2094: Q\_n1\_b0 & 1.0e-04 & 1.0e-04 \\
2095: Q\_n1\_b0.5 & 0.50 & 0.50 \\
2096: Q\_n1\_b1 & 1.00 & 1.00 \\
2097: Q\_n1\_b2 & 2.00 & 2.00 \\
2098: Q\_n1\_be & 3.23 & 3.23 \\
2099: Q\_n1\_b4 & 4.00 & 4.00 \\
2100: Q\_n1\_b5 & 5.00 & 5.00 \\
2101: Q\_n1\_b10 & 10.0 & 10.0 \\
2102: \cutinhead{solar abundance, n=0.01, v=100,125,...1000}
2103: T\_n0.01\_b0.001 & 0.001 & 0.010\\
2104: T\_n0.01\_b0.01 & 0.010 & 0.10 \\
2105: T\_n0.01\_b0.05 & 0.050 & 0.50 \\
2106: T\_n0.01\_b0.1 & 0.10 & 1.00 \\
2107: T\_n0.01\_b0.2 & 0.20 & 2.00 \\
2108: T\_n0.01\_b0.4 & 0.40 & 4.00 \\
2109: T\_n0.01\_b0.5 & 0.50 & 5.00 \\
2110: T\_n0.01\_b1 & 1.0 & 10.0 \\
2111: T\_n0.01\_b10 & 10.0 & 100 \\
2112: \cutinhead{solar abundance, n=0.1, v=100,125,...1000}
2113: U\_n0.1\_b0.0001 & 1.0e-04 & 0.000316 \\
2114: U\_n0.1\_b0.001 & 0.001 & 0.00316 \\
2115: U\_n0.1\_b0.01 & 0.01 & 0.0316 \\
2116: U\_n0.1\_b0.05 & 0.05 & 0.158 \\
2117: U\_n0.1\_b0.1 & 0.10 & 0.316 \\
2118: U\_n0.1\_b0.2 & 0.20 & 0.632 \\
2119: U\_n0.1\_b0.32 & 0.32 & 1.01 \\
2120: U\_n0.1\_b0.4 & 0.40 & 1.26 \\
2121: U\_n0.1\_b0.5 & 0.50 & 1.58 \\
2122: U\_n0.1\_b0.632 & 0.63 & 2.00 \\
2123: U\_n0.1\_b1.0 & 1.00 & 3.16 \\
2124: U\_n0.1\_b1.26 & 1.26 & 4.00 \\
2125: U\_n0.1\_b1.58 & 1.58 & 5.00 \\
2126: U\_n0.1\_b2.0 & 2.00 & 6.32 \\
2127: U\_n0.1\_b3.16 & 3.16 & 10.0 \\
2128: U\_n0.1\_b4.0 & 4.00 & 12.6 \\
2129: U\_n0.1\_b5.0 & 5.00 & 15.8 \\
2130: U\_n0.1\_b10 & 10.0 & 31.6 \\
2131: \cutinhead{solar abundance, n=10.0, v=100,125,...1000}
2132: V\_n10\_b0.001 & 0.001 & 0.000316 \\
2133: V\_n10\_b0.01 & 0.01 & 0.00316 \\
2134: V\_n10\_b0.1 & 0.10 & 0.0316 \\
2135: V\_n10\_b1 & 1.00 & 0.316 \\
2136: V\_n10\_b1.58 & 1.58 & 0.500 \\
2137: V\_n10\_b3.16 & 3.16 & 1.00 \\
2138: V\_n10\_b5 & 5.00 & 1.58 \\
2139: V\_n10\_b6.32 & 6.32 & 2.00 \\
2140: V\_n10\_b10 & 10.0 & 3.16 \\
2141: V\_n10\_b10.2 & 10.2 & 3.23 \\
2142: V\_n10\_b12.65 & 12.6 & 4.00 \\
2143: V\_n10\_b15.8 & 15.8 & 5.00 \\
2144: V\_n10\_b20 & 20.0 & 6.32 \\
2145: V\_n10\_b30 & 30.0 & 9.49 \\
2146: V\_n10\_b40 & 40.0 & 12.6 \\
2147: V\_n10\_b50 & 50.0 & 15.8 \\
2148: V\_n10\_b100 & 100 & 31.6 \\
2149: \cutinhead{solar abundance, n=100, v=100,125,...1000}
2150: L\_n100\_b0.001 & 0.001 & 0.0001 \\
2151: L\_n100\_b0.01 & 0.01 & 0.001 \\
2152: L\_n100\_b0.1 & 0.10 & 0.010 \\
2153: L\_n100\_b1 & 1.00 & 0.10 \\
2154: L\_n100\_be & 32.3 & 3.23 \\
2155: L\_n100\_b5 & 5.0 & 0.50 \\
2156: L\_n100\_b10 & 10.0 & 1.00 \\
2157: L\_n100\_b20 & 20.0 & 2.00 \\
2158: L\_n100\_b40 & 40.0 & 4.00 \\
2159: L\_n100\_b50 & 50.0 & 5.00 \\
2160: L\_n100\_b100 & 100 & 10.0 \\
2161: \cutinhead{solar abundance, n=1000, v=100,125,...1000}
2162: S\_n1000\_b0.01 & 0.01 & 0.000316 \\
2163: S\_n1000\_b0.1 & 0.1 & 0.00316 \\
2164: S\_n1000\_b1 & 1.0 & 0.0316 \\
2165: S\_n1000\_b5 & 5.00 & 0.158 \\
2166: S\_n1000\_b10 & 10.0 & 0.316 \\
2167: S\_n1000\_b16 & 16.0 & 0.51 \\
2168: S\_n1000\_b32 & 32.00 & 1.01 \\
2169: S\_n1000\_b63 & 63.00 & 2.00 \\
2170: S\_n1000\_b100 & 100.0 & 3.16 \\
2171: S\_n1000\_b126 & 126.0 & 4.00 \\
2172: S\_n1000\_b160 & 160.0 & 5.06 \\
2173: S\_n1000\_b316 & 316.0 & 10.0 \\
2174: S\_n1000\_b1000 & 1000 & 31.6 \\
2175: \enddata
2176: \end{deluxetable}
2177:
2178: \clearpage
2179: \LongTables
2180: \begin{landscape}
2181: \include{tab4}
2182: \end{landscape}
2183:
2184: \clearpage
2185: \LongTables
2186: \begin{landscape}
2187: \include{tab5}
2188: \end{landscape}
2189:
2190: \LongTables
2191: \include{tab6}
2192:
2193: \LongTables
2194: \include{tab7}
2195:
2196: \LongTables
2197: \include{tab8}
2198:
2199: \end{document}
2200: